Issue 131: the trigger cannot be one build per changed module

Correcting my own framing. The modules are a graph, not a flat set: several build
from one repository, and some are the base another compiles on. Firing a build per
changed module starts work that cannot succeed yet and reports one cause as a
failure per dependent.

Observed catching the mesh up by hand today — a compiler base had to move first,
and when it failed the right behaviour was for its fifteen dependents to wait, not
to fail identically. So what reacts to the forge's event has to resolve what changed
into an order and hold a dependent while its base is unbuilt. That is a bigger thing
than "rebuild what the commit touched", and cheaper to know now.

.claude/skills/hq-defer/SKILL.md

@@ -0,0 +1,78 @@
+---
+name: hq-defer
+description: Use when a thought is raised that should be remembered but NOT worked on now — an aside during other work, a "we should look at X someday", a known gap nobody is assigning yet. Triggers on "defer this", "park this", "register this thought", "note this for later", "don't work on it, just remember it". Records and returns to whatever was already in progress.
+---
+
+# hq-defer
+
+Parks a thought so it is not lost, **without moving the work off course.** The defer is the
+point: the thought is recorded and the previous task resumes.
+
+This skill wraps no playbook, because deferring is not part of the development cycle — it is
+what happens *before* something enters it. A parked thought has no number, no owner and no
+status, and that is correct.
+
+## Where it goes, and why not the repository
+
+Record it as **one memory file** in Claude's persistent memory directory for this project (the
+path is given in the session's memory instructions), with `metadata.type: project`, plus a
+one-line pointer in `MEMORY.md`.
+
+**Not** in `04-ISSUES`, `01-RESEARCH` or anywhere else in the repository:
+
+- A parked thought is not an issue or a research effort. Giving it a number asserts it has been
+  triaged, which is exactly what deferring says has not happened.
+- A shared "deferred" or "someday" document is a **central status file**, which
+  [`AGENTS.md`](../../../AGENTS.md) forbids. Status lives in frontmatter on real records, and a
+  parked thought has no real record yet.
+- A repository write means a branch, a commit and a pull request — drift, which is the one thing
+  this skill exists to avoid.
+
+## Steps
+
+1. Write the memory file. Slug is kebab-case and descriptive of the thought, not of the act of
+   deferring.
+
+   ```markdown
+   ---
+   name: <kebab-slug>
+   description: Deferred note — <one line>
+   metadata:
+     type: project
+   ---
+
+   Raised and deliberately deferred on YYYY-MM-DD: **<the thought, in the user's own terms>**
+
+   **Why:** what was being worked on when it came up, and that deferring was intentional so
+   that work was not pulled off course.
+
+   **How to apply:** treat as an open thread, not an assignment. Do not start on it
+   unprompted. If it graduates it needs an HQ home first — an issue under playbook
+   [03](../../../00-META/process/03-issues.md) if a stated behaviour does not happen, or
+   research under playbook [01](../../../00-META/process/01-research.md) if it is still an
+   idea. Say which is undetermined, if it is.
+   ```
+
+2. Append one line to `MEMORY.md`: `- [<Title>](<kebab-slug>.md) — deferred YYYY-MM-DD; parked, no HQ
+   record, do not start unprompted`.
+3. Convert relative dates to absolute before writing. "Last week" is worthless in six months.
+4. Check for an existing memory covering the same thought and update it instead of adding a
+   duplicate.
+
+## Then stop
+
+Reply in **at most two lines** — what was recorded, and that it is parked — and **return to
+whatever was in progress before.** Do not summarise the parked thought back at length, do not
+propose a plan for it, do not ask which playbook it belongs to, and do not open anything.
+
+If nothing was in progress, say only that it is recorded.
+
+## Do not
+
+- Do not create an issue, a research effort, a decision record or a design document.
+- Do not create a branch, commit or pull request.
+- Do not start investigating the thought, however cheap the first check looks.
+- Do not name nodes, domains, addresses, absolute paths or usernames in the memory file — the
+  thought may later be quoted into this repository, which is public.
+- Do not decide whether it is an issue or research when the evidence does not say. Recording
+  "undetermined" is the honest outcome and costs nothing later.

00-META/README.md

@@ -9,6 +9,7 @@ works — plus the engineering practice that holds across everything Novox build
 | [`context.md`](context.md) | The environment — conditions, not aspirations |
 | [`effect.md`](effect.md) | What is different when the work is done |
 | [`how-we-build.md`](how-we-build.md) | The rules that hold across the mesh, each one earned. **The source of the mesh constitution** — the governed page the mesh injects into design sessions is derived from it. |
+| [`glossary.md`](glossary.md) | One name per thing — the authority on vocabulary, and the words that were retired |
 | [`repos.md`](repos.md) | Where implementation lives, and what each repository owns |
 | [`process/`](process/) | The playbooks — how work moves through this repository, for engineers and agents alike |
 

00-META/checks/README.md

@@ -26,6 +26,7 @@ indistinguishable from one that cannot.
 | `numbering` | the number in the filename is the number in the heading | — |
 | `topics` | every record names a topic the index knows | — |
 | *(index.py)* | the written reading order matches what the records say | — |
+| `status-vs-code` | a to-be document naming specific code is not still `designed` | **ten documents**, several with a *What was built* section, describing lab-proven code |
 
 ## What is deliberately not checked
 
@@ -47,3 +48,10 @@ what to do.
 
 State what incident it would have caught, and make it fail before you make it pass. A check
 whose failure has never been observed is a guess about its own correctness.
+
+## cycle.py
+
+The development cycle, checked ([ADR 0080](../../02-DECISIONS/0080-the-development-cycle-is-checked.md)):
+a to-be design names a decision, an in-progress/implemented design names its owning code, a
+located/fixed issue names its owner, a fixed/resolved issue says what fixed it, a graduated
+research overview says what it became. `python3 00-META/checks/cycle.py`

00-META/checks/cycle.py

@@ -0,0 +1,188 @@
+#!/usr/bin/env python3
+"""The development cycle, checked.
+
+The knowledge flow (00-META/process/00-overview.md) says work moves idea -> research ->
+decision -> to-be design -> code, and symptom -> issue -> diagnosis -> fix. Those are rules,
+and a rule states how it is checked (AGENTS.md) -- this is how. Everything here reads only
+frontmatter, because status lives in frontmatter and nowhere else.
+
+What is enforced:
+
+  design  every 03-DESIGN doc parses, carries `layer:` matching its directory, and a
+          known `status:`. A TO-BE doc names at least one decision (`decisions:`) -- no
+          design without a decision -- and once `in-progress` or `implemented` it names
+          its owning code (`code:`) -- no development without a design that says where.
+  issues  a known `status:`; once `located`, `located-in:` names the owner;
+          once `resolved`, `fixed-by:` says what fixed it (prose counts --
+          "nothing, the capability existed" is an answer).
+  research a known `status:`; a `graduated` overview says what it `became:`, and every
+          target it names exists.
+  decisions every accepted record is REACHABLE from the cycle: cited by a design doc's
+          frontmatter, a research overview, an issue report, a 00-META document, or another
+          record's extends/supersedes chain. A decision nothing points at is one nobody will
+          find by following pointers -- which is how records go stale in people's heads.
+
+Deliberately NOT enforced: `resolved` issues may leave `located-in` empty (a symptom that
+turned out not to be a defect has no owner), and as-is docs need no decisions (they
+describe what exists, not what was decided).
+
+    python3 00-META/checks/cycle.py
+"""
+
+import glob
+import os
+import re
+import sys
+
+ROOT = os.path.normpath(os.path.join(os.path.dirname(__file__), "..", ".."))
+
+DESIGN_STATUSES = {"proposed", "designed", "in-progress", "implemented", "abandoned"}
+ISSUE_STATUSES = {"open", "diagnosing", "located", "resolved", "wontfix"}
+RESEARCH_STATUSES = {"active", "graduated", "abandoned"}
+
+
+def rel(path):
+    return os.path.relpath(path, ROOT)
+
+
+def frontmatter(path):
+    """The YAML block between the first two --- lines, as {key: raw-value-string}.
+
+    Minimal on purpose, like records.py: enough for the fields these checks read. A list
+    value (block or inline) is joined into its items; a scalar stays a string.
+    """
+    text = open(path, encoding="utf-8").read()
+    m = re.match(r"^---\n(.*?)\n---", text, re.S)
+    if not m:
+        return None
+    front, out, key = m.group(1), {}, None
+    for line in front.split("\n"):
+        item = re.match(r"^\s+-\s*(.+?)\s*$", line)
+        if item and key:
+            out[key].append(item.group(1))
+            continue
+        kv = re.match(r"^([A-Za-z-]+):\s*(.*)$", line)
+        if not kv:
+            continue
+        key, value = kv.group(1), kv.group(2).strip()
+        if value.startswith("[") and value.endswith("]"):
+            out[key] = [v.strip() for v in value[1:-1].split(",") if v.strip()]
+        elif value == "":
+            out[key] = []  # a block list may follow; stays [] if nothing does
+        else:
+            out[key] = value
+    return out
+
+
+def listy(front, key):
+    v = front.get(key)
+    if v is None:
+        return []
+    return v if isinstance(v, list) else ([v] if str(v).strip() else [])
+
+
+def main():
+    failures = []
+
+    def bad(path, why):
+        failures.append("  %s: %s" % (rel(path), why))
+
+    # ---- design ------------------------------------------------------------------------
+    for layer, name in (("00-as-is", "as-is"), ("01-to-be", "to-be")):
+        for path in sorted(glob.glob(os.path.join(ROOT, "03-DESIGN", layer, "*.md"))):
+            if os.path.basename(path) == "README.md":
+                continue
+            front = frontmatter(path)
+            if front is None:
+                bad(path, "no frontmatter")
+                continue
+            if front.get("layer") != name:
+                bad(path, "layer is %r; this directory is %s" % (front.get("layer"), name))
+            status = front.get("status")
+            if status not in DESIGN_STATUSES:
+                bad(path, "status %r is not one of %s" % (status, sorted(DESIGN_STATUSES)))
+            if name == "to-be":
+                if not listy(front, "decisions"):
+                    bad(path, "names no decisions -- no design without a decision")
+                if status in ("in-progress", "implemented") and not listy(front, "code"):
+                    bad(path, "status %s but code: names no owner -- no development "
+                              "without a design that says where" % status)
+
+    # ---- issues ------------------------------------------------------------------------
+    for path in sorted(glob.glob(os.path.join(ROOT, "04-ISSUES", "*", "00-report.md"))):
+        front = frontmatter(path)
+        if front is None:
+            bad(path, "no frontmatter")
+            continue
+        status = front.get("status")
+        if status not in ISSUE_STATUSES:
+            bad(path, "status %r is not one of %s" % (status, sorted(ISSUE_STATUSES)))
+        if status in ("located", "resolved") and not listy(front, "located-in"):
+            bad(path, "status %s but located-in is empty" % status)
+        if status == "resolved" and not listy(front, "fixed-by"):
+            bad(path, "status %s but fixed-by says nothing" % status)
+
+    # ---- research ----------------------------------------------------------------------
+    for path in sorted(glob.glob(os.path.join(ROOT, "01-RESEARCH", "*", "00-overview.md"))):
+        front = frontmatter(path)
+        if front is None:
+            bad(path, "no frontmatter")
+            continue
+        status = front.get("status")
+        if status not in RESEARCH_STATUSES:
+            bad(path, "status %r is not one of %s" % (status, sorted(RESEARCH_STATUSES)))
+        if status == "graduated":
+            became = listy(front, "became")
+            if not became:
+                bad(path, "graduated but became: names nothing")
+            for target in became:
+                if not os.path.exists(os.path.join(ROOT, target)):
+                    bad(path, "became names %s, which does not exist" % target)
+
+    # ---- decisions -------------------------------------------------------------------
+    records = {}
+    for path in sorted(glob.glob(os.path.join(ROOT, "02-DECISIONS", "[0-9]*.md"))):
+        front = frontmatter(path)
+        records[os.path.basename(path)] = (path, (front or {}).get("status"))
+
+    cited = set()
+    sources = (glob.glob(os.path.join(ROOT, "03-DESIGN", "*", "*.md"))
+               + glob.glob(os.path.join(ROOT, "01-RESEARCH", "*", "00-overview.md"))
+               + glob.glob(os.path.join(ROOT, "04-ISSUES", "*", "00-report.md"))
+               + glob.glob(os.path.join(ROOT, "00-META", "**", "*.md"), recursive=True))
+    for path in sources:
+        text = open(path, encoding="utf-8").read()
+        if os.sep + "03-DESIGN" + os.sep in path:
+            # A design doc's governing citations live in frontmatter; a prose mention is
+            # commentary, not a home.
+            m = re.match(r"^---\n(.*?)\n---", text, re.S)
+            text = m.group(1) if m else ""
+        for m in re.finditer(r"([0-9]{4}-[^\s\)\],#]+\.md)", text):
+            cited.add(os.path.basename(m.group(1)))
+    for name in records:
+        front = frontmatter(records[name][0]) or {}
+        for key in ("extends", "supersedes", "superseded-by"):
+            v = front.get(key)
+            if isinstance(v, str) and v:
+                cited.add(os.path.basename(v))
+    for name, (path, status) in records.items():
+        if status == "accepted" and name not in cited:
+            bad(path, "an accepted decision nothing in the cycle cites -- give it a home in a "
+                      "design doc's decisions:, a research became:, an issue, or 00-META")
+
+    checked = (
+        len(glob.glob(os.path.join(ROOT, "03-DESIGN", "0*", "*.md")))
+        + len(glob.glob(os.path.join(ROOT, "04-ISSUES", "*", "00-report.md")))
+        + len(glob.glob(os.path.join(ROOT, "01-RESEARCH", "*", "00-overview.md")))
+        + len(records)
+    )
+    if failures:
+        print("cycle: %d document(s) break the development cycle:" % len(failures))
+        print("\n".join(failures))
+        return 1
+    print("cycle: %d documents checked, the chain holds" % checked)
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())

00-META/checks/records.py

@@ -284,6 +284,107 @@ def check_numbering(failures, records):
             )
 
 
+def check_progressive_insights(failures, records):
+    """A correction made inside a record is marked and dated, or it is a silent rewrite.
+
+    A record may be corrected in place when a *fact* in it went stale and the decision still
+    stands (`02-DECISIONS/README.md`, "Progressive insight"). The whole safety of that allowance
+    is that the correction is legible in the record rather than only in a diff nobody reads, so
+    the form is what is checked here: every mention of an insight is the marker, the marker
+    carries an ISO date, and that date is not earlier than the decision's own — an insight
+    predating the decision it corrects is a copied marker, not a correction.
+
+    What this cannot check is an edit made with no marker at all. Nothing mechanical can; that
+    one is the reviewer's, reading the diff. The check keeps the *marked* path honest so that an
+    unmarked change stands out as the anomaly it is.
+    """
+    phrase = re.compile(r"progressive insight", re.I)
+    # Both patterns stay on one line: a bold run does not span paragraphs, and `[^*]*` across
+    # newlines will happily join an unrelated `**` far above to the marker below, reporting the
+    # whole span between them. It did exactly that the first time this ran.
+    # Trailing words after the date are allowed — "— 2026-09-26, correcting the one above." — so
+    # an insight can say what it relates to. Only the date's presence and position are fixed.
+    marker = re.compile(r"\*\*Progressive insights?[ \t]*[\u2014\u2013-][ \t]*(\d{4}-\d{2}-\d{2})[^*\n]*\*\*")
+    loose = re.compile(r"\*\*[^*\n]*[Pp]rogressive insights?[^*\n]*\*\*")
+    iso = re.compile(r"^\d{4}-\d{2}-\d{2}$")
+
+    for number, record in sorted(records.items()):
+        text = record["text"]
+        if not phrase.search(text):
+            continue
+        decided = str(record["front"].get("date", ""))
+        good = [(m.start(), m.end(), m.group(1)) for m in marker.finditer(text)]
+
+        for m in loose.finditer(text):
+            if any(s <= m.start() and m.end() <= e for s, e, _ in good):
+                continue
+            # A bold run carrying a link is discussing an insight — usually another record's —
+            # rather than marking one. A marker never needs to cite anything.
+            if "](" in m.group(0):
+                continue
+            failures.add("insights", rel(record["path"]),
+                         "a progressive insight is not in the dated marked form "
+                         "'**Progressive insight \u2014 YYYY-MM-DD.**': %s" % m.group(0))
+
+        for _, _, stamp in good:
+            if decided and iso.match(decided) and stamp < decided:
+                failures.add("insights", rel(record["path"]),
+                             "a progressive insight dated %s predates the decision (%s)"
+                             % (stamp, decided))
+
+        covered = [(s, e) for s, e, _ in good]
+        for m in phrase.finditer(text):
+            if any(s <= m.start() and m.end() <= e for s, e in covered):
+                continue
+            line = text.rfind("\n", 0, m.start()) + 1
+            end = text.find("\n", m.end())
+            whole = text[line:end if end != -1 else len(text)]
+            if whole.lstrip().startswith("#"):
+                continue
+            # A line that also carries a link is discussing the rule, not marking a correction:
+            # a marker never needs to cite anything, and a record that reasons about the policy
+            # must be able to name it. Bare prose with no citation is the informal marking this
+            # is here to catch.
+            if "](" in whole:
+                continue
+            if loose.search(text, line, text.find("\n", m.end()) + 1 or len(text)):
+                continue
+            failures.add("insights", rel(record["path"]),
+                         "'progressive insight' appears unmarked; a correction is marked and "
+                         "dated, or it is a silent rewrite")
+
+
+def check_status_against_code(failures):
+    """A design document naming specific code may not still call itself `designed`.
+
+    **Naming a file is a claim that the file implements this**, so the two fields have to agree.
+    They drifted: ten to-be documents named working, lab-proven code — several with a *What was
+    built* or *Raised, and observed* section — while still saying nothing had been built.
+
+    Deliberately weak, and that is the point of it being mechanical. It cannot tell whether the
+    prose is true, only that a document has stopped claiming to be unbuilt once it points at
+    something. `code: [mesh-controller]` — a repository with no path — is a plan and stays
+    `designed`.
+    """
+    for path in markdown_files():
+        if not rel(path).startswith("03-DESIGN/01-to-be/") or path.endswith("README.md"):
+            continue
+        front = frontmatter(read(path))
+        if front.get("status") != "designed":
+            continue
+        for entry in front.get("code") or []:
+            named = re.sub(r"\s*\(.*\)$", "", entry).strip().split(None, 1)
+            if len(named) > 1:
+                failures.add(
+                    "status-vs-code",
+                    rel(path),
+                    f"`designed`, but names {named[1]!r} in {named[0]}. Naming a file claims "
+                    f"it implements this — use `in-progress`, or `implemented` once it is "
+                    f"defensible from that repository's main branch.",
+                )
+                break
+
+
 def main():
     failures = Failures()
     records = load_records()
@@ -293,6 +394,8 @@ def main():
     check_supersession_symmetry(failures, records)
     check_numbering(failures, records)
     check_topics(failures, records)
+    check_status_against_code(failures)
+    check_progressive_insights(failures, records)
     print(f"records: {len(records)} decision records checked")
     return failures.report()
 

00-META/context.md

@@ -30,7 +30,7 @@ named, and nothing should be designed around a particular one existing.
 - **A hosted model provider** supplies the thinking for non-human agents, drawn from a
   shared pool of subscriptions — which is why budget pacing is a first-class concern.
 - **Long-lived user services** rather than an orchestrator. No cluster scheduler, no cloud
-  control plane.
+  controller.
 
 Defaults, not mandates. A second model provider is anticipated by design; nothing in the
 domain may assume one vendor's credential lifecycle.

00-META/glossary.md

@@ -0,0 +1,82 @@
+# Glossary — the words this repository uses, and the ones it stopped using
+
+One name per thing. This page is the authority; where an older record says something else, that
+record is being superseded, not this page. It exists because the terms kept drifting in
+conversation — control plane / controller / master / hub for one thing, substrate / foundation for
+another — and a mesh you cannot name precisely is a mesh two people describe differently.
+
+## The mesh and its machines
+
+- **node** — a machine in the mesh. There are 0..n of them, and each runs the host agent. A node is
+  just a machine that has joined; being one implies nothing about what it runs.
+- **control-node** — the one node that also holds the `mesh-controller` seat. There is exactly one
+  per mesh. "control-node" is not a separate kind of machine — it is a node that additionally runs
+  the controller (and, today, the foundation). Lose it and the other nodes keep running what they
+  were last told; they simply cannot be told anything new.
+- ~~master / slave~~, ~~hub / peer~~ — not used. The relationship is *controller and nodes*, and no
+  node is subordinate: a node applies declarations on its own and survives the control-node dying.
+
+## What runs the mesh
+
+- **controller** — the component that decides what each node should be, holds the mesh's records,
+  and tells nodes over the broker. Replaces **"control plane"** (borrowed from networking's
+  control-plane/data-plane, and opaque here).
+- **mesh-controller** — the module that runs the controller. It **claims** the `mesh-controller`
+  seat at mesh scope, which is what makes it singular. Replaces the module name **`mesh-control`**.
+  (The git repository has been renamed `mesh-control` -> `mesh-controller` on the forge; the module,
+  container and image it produces are `mesh-controller`.)
+- **foundation** — the store and the broker, raised at genesis before any module system exists.
+  Replaces **"substrate"** (a biology metaphor that landed for no one). The foundation is not a
+  third thing beside the store and broker — it *is* those two, named together.
+- **store** — the one postgres server. It holds the controller's own context databases
+  (`inventory`, `identity`, `licences` — a context owns its store, [ADR 0008](../02-DECISIONS/0008-a-context-owns-its-store.md))
+  and every module's own database. One server, many databases — never one shared "mesh database".
+- **bus** — the mesh's own nervous system: NATS, one per mesh, carrying every link the mesh has —
+  control, declarations, builds, events, tool calls
+  ([ADR 0106](../02-DECISIONS/0106-the-bus-is-nats.md)). A module reaches it by requiring
+  `mesh-bus` ([ADR 0128](../02-DECISIONS/0128-the-mesh-bus-is-required-not-ambient.md)); one that
+  does not require it has no account on it. Held by the `mesh-broker` seat, which is named after
+  the *role* rather than the server, so the server can change without the seat doing so.
+- **the deprecated broker** — the lavinmq module. It was the mesh's bus and is not any more. It
+  keeps running as an **ordinary provider** of the `amqp` provision, for modules that need a
+  message broker of their own the way something needs a database
+  ([ADR 0127](../02-DECISIONS/0127-amqp-is-a-provision-not-the-bus.md)) — no seat, not foundation,
+  never raised at genesis, and a mesh that never installs it is complete.
+
+  Say *the deprecated broker*, not "the compatibility broker" (it serves the mesh's own modules,
+  not only the predecessor's) and not "the AMQP broker" (naming it after a protocol invites
+  describing the bus by contrast with it, which is backwards: the bus is the mesh's nervous
+  system and this is a module).
+
+## What the mesh stores and serves
+
+- **package** — what code resolves when it is **compiled**: an npm/cargo/pypi dependency, by
+  **version**. Served by the **package-registry** (gitea). Only a builder talks to it.
+- **artifact** — what the mesh delivers to a machine to **install and run**: an OCI image, by
+  **digest**. Served by the **artifact-store** (distribution). Every node pulls from it.
+- These are two protocols, not one store being weak — see [ADR 0075](../02-DECISIONS/0075-two-stores-and-which-provides-what.md).
+
+## How modules relate to the mesh
+
+- **seat** — a named role at a scope (node / site / mesh), held by a module assignment, from a
+  **closed set** the mesh defines: a claim naming a seat outside the set is refused. A seat may
+  **deliver a provision**, and its holder is then the mesh's answer for it when several modules
+  provide it ([ADR 0126](../02-DECISIONS/0126-a-module-declares-its-own-seats.md) (superseding [ADR 0110](../02-DECISIONS/0110-a-seat-is-a-module-assignment-from-a-closed-set.md))).
+  The set, with who holds each seat, is the overview of what a mesh has
+  ([26 — The seats](../03-DESIGN/01-to-be/26-the-seats.md)). A seat has a **capacity**: a
+  capacity-1 seat is exclusive (one holder); a higher-capacity seat is a **bench** (several holders
+  coexist).
+- **claim** — a module taking a spot on a seat. `claims: [{name, scope}]` in a manifest. A
+  mesh-scoped exclusive claim is how the mesh says "there is one of me". A foundation seat is
+  named after the server it guards: the `mesh-controller`, `postgres` and `lavinmq` modules claim
+  the `mesh-controller`, `mesh-store` and `mesh-broker` seats ([ADR 0079](../02-DECISIONS/0079-the-foundation-seats-are-named-after-their-servers.md)).
+- **provision** — a service one module `provides` and others `require`; the mesh resolves a provider
+  and wires the two with an endpoint and a credential. A provision is a service you offer, a seat
+  is a role you occupy, and the two meet where a seat delivers a provision: occupying the seat is
+  what makes a module *the* provider of it.
+
+## How this page is kept
+
+A new name for an existing thing lands here first, in the same change that introduces it in code. A
+record under `02-DECISIONS/` keeps whatever word it was written with — those are immutable — so a
+term retired here may still appear there, and the mapping above is how to read it.

00-META/process/00-overview.md

@@ -1,6 +1,10 @@
 # Process — overview
 
-How work moves through HQ, and who may do what. Every other document in this folder is a
+How work moves through HQ, and who may do what. In industry terms this is **spec-driven
+development, with provenance**: the decision is the why, the design doc is the spec, `code:`
+names the implementation, and the lab beds are the conformance tests — and unlike the common
+form, the chain itself is checked ([ADR 0080](../../02-DECISIONS/0080-the-development-cycle-is-checked.md),
+[0081](../../02-DECISIONS/0081-a-decision-nothing-cites-is-not-yet-in-the-chain.md)). Every other document in this folder is a
 playbook: trigger, who runs it, steps, outputs. Engineers and agents follow the same
 playbooks; agents must not act outside them.
 
@@ -49,6 +53,19 @@ the expensive half.
 | [03](03-issues.md) | Issues | Something is wrong — often with the owner unknown |
 | [04](04-build-handoff.md) | Build handoff | A design is ready to be built |
 | [05](05-constitution-sync.md) | Constitution sync | `how-we-build.md` changed a rule the mesh enforces |
+| [06](06-writing-a-module.md) | Writing a module | Something that runs today must run on the mesh |
+| [07](07-feature-branches.md) | Feature branches across repos | Work that changes code, in one repo or several at once |
+
+## The cycle is checked
+
+The flow above is a rule, and a rule states how it is checked:
+[`00-META/checks/cycle.py`](../checks/cycle.py) refuses a to-be design that names no
+decision, an `in-progress`/`implemented` design that names no owning code, an issue marked
+`located`/`fixed` with no owner or `fixed`/`resolved` with no fix, and a `graduated`
+research overview that does not say what it became. Run it with `records.py` and `index.py`
+before any HQ merge. What the checks cannot see — that code work actually started from a
+handoff — is held by playbooks [04](04-build-handoff.md) and [07](07-feature-branches.md):
+a feature branch exists because a design or an issue sent it.
 
 ## Status lives in frontmatter
 

00-META/process/02-graduation.md

@@ -33,7 +33,10 @@
 A design changes only through a decision.
 
 1. Write the decision record. If it reverses an earlier one, the earlier record's `status:`
-   becomes `superseded-by: 02-DECISIONS/NNNN-....md` — **its text is never edited**.
+   becomes `superseded-by: 02-DECISIONS/NNNN-....md` — **its reasoning is never rewritten**. If the
+   earlier record is sound and only a *fact* in it went stale, that is a **progressive insight**,
+   corrected in place and marked in the record rather than superseded
+   ([`02-DECISIONS/README.md`](../../02-DECISIONS/README.md)).
 2. Edit the to-be design document and set `updated:` to today.
 3. If the amendment came from an issue, set that issue's `amended-design:` to the document
    path.
@@ -54,4 +57,5 @@ Implementation state is a third axis, independent of both design and decision.
 
 - Do not move a to-be document into `00-as-is/`. Write the as-is document; both stand.
 - Do not edit an as-is document to describe an intention. That is what the to-be layer is for.
-- Do not change a decision record's meaning. Supersede it.
+- Do not change a decision record's meaning. Supersede it. Correcting a fact it got wrong, while
+  the decision stands, is a progressive insight — marked and dated in the record, never silent.

00-META/process/05-constitution-sync.md

@@ -1,5 +1,8 @@
 # Playbook 05 — Constitution sync
 
+Implements [ADR 0021](../../02-DECISIONS/0021-hq-is-the-source-of-the-constitution.md): HQ is
+the source of the constitution, and the knowledge-base page is derived, never edited.
+
 **Trigger.** [`how-we-build.md`](../how-we-build.md) changed a rule that the mesh enforces at
 runtime.
 

00-META/process/06-writing-a-module.md

@@ -0,0 +1,107 @@
+# Playbook 06 — Writing a module
+
+**Trigger.** Something that runs today must run on the mesh, or a new capability must be
+declarable.
+
+**Who runs it.** Whoever is porting or writing it.
+
+*Written 2026-09-01 from doing this for the first time end to end. Every step below exists
+because skipping it cost something.*
+
+## Before anything: read what runs
+
+**A module is written from the thing, not from memory of the thing.** For a port, that means its
+current compose file, its environment, and where its data actually sits. Assumptions about any of
+the three have been wrong every time they were not checked.
+
+Three questions, answered from the machine:
+
+| | why it decides something |
+|---|---|
+| **what containers, and how do they find each other?** | more than one means a `network`; names between them must match what the software is configured to dial |
+| **where is its data?** | a bind mount moves with a path; a named volume does not; an anonymous volume is already losing data on every redeploy |
+| **which values are secret, and which are merely settings?** | a secret goes in `own-secrets` or a grant; a setting goes in the manifest and may be overridden per node |
+
+## The steps
+
+1. **Name what it provides and requires**, if anything. A name is what a consumer is coupled to,
+   not the role it plays ([ADR 0027](../../02-DECISIONS/0027-a-provision-names-what-the-consumer-is-coupled-to.md)):
+   `postgres-database`, not `database`. Most modules provide nothing and require nothing — an
+   application is usually a leaf.
+
+2. **Declare capabilities, not dependencies, for facts about the machine.** `container-runtime`,
+   `package-manager`, `seat`. A capability is detected and refused against; it is not something a
+   module can install.
+
+3. **Write the resources in the order they must happen.** They are applied in the order written
+   and orphans are removed in reverse, so a `network` is written before the containers that join
+   it and removed after them.
+
+4. **Put every secret in a file, never in `env`.** A declaration travels over the broker in plain
+   text: a password in `env` is a password the broker sees. **The mesh delivers parts; a module
+   that needs them combined combines them.**
+
+   **A sealed file holds the password and nothing else** — no key, no `=`, no newline that means
+   anything. So `env-file` must never point at one. It points at a file the module *declares*,
+   whose content leaves a hole:
+
+   ```
+   own-secrets   superuser → /var/lib/postgres/superuser.secret     the password, alone
+   a file        /var/lib/postgres/superuser.env, mode 0600,
+                 content: POSTGRES_PASSWORD=${secret:superuser}
+   the container env-file: [/var/lib/postgres/superuser.env]
+   ```
+
+   The host fills the hole on the machine, which is the only place both halves exist — the mesh
+   discarded the value
+   ([credentials and their rotation](../../03-DESIGN/01-to-be/13-credentials-and-their-rotation.md)).
+   A **provisioner** is the exception: it reads a password file, so it mounts the `.secret`
+   directly.
+
+   Every example module in `mesh-controller` had this wrong and shipped: `own-secrets` pointing at a
+   path *named* `.env`, mounted as `env-file`, holding a bare password. Docker reads that as a
+   malformed line and the container starts **with no password set at all** — not a failure to
+   start, a service running on the wrong credential. They parsed and they resolved. Two tests in
+   `examples/modules` now refuse both halves of it.
+
+   Add `restart-on` naming the env file, or the container keeps the credential it started with
+   through every rotation.
+
+5. **Pin every image by digest.** A tag moves. The manifest in a repository names artifacts; the
+   manifest the mesh holds names digests, and they are not the same document.
+
+6. **Decide generate or accept.** A new module's credential is generated. **An adopted one keeps
+   the credential it already has** — `secret accept` — because minting a new password for a
+   database that already exists locks the application out of its own data.
+
+7. **Add a provisioner only if the software cannot read a file.** A proxy that watches a
+   directory needs nothing. PostgreSQL needs `CREATE ROLE`, an object store needs a bucket and a
+   policy, an identity provider needs a realm and a client — those need a small program beside
+   them. It reads what the mesh granted and reconciles; it does not decide anything.
+
+8. **Prove it in the lab, against the real software.** Not that a container started — that the
+   thing works: the credential authenticates, a wrong one is refused, the containers reach each
+   other, the data survives a restart.
+
+## What the first port actually cost
+
+Six attempts, one real bug. Recorded because the ratio is the lesson: **the mesh was right every
+time and the scaffolding was not.**
+
+- A shape existed in the language and no host implemented it, so every declaration carrying one
+  was refused whole — correctly, and the host said exactly that. **Nobody was reading the host's
+  log.** Read it first; it is the only place that says why a machine did nothing.
+- A blind find-and-replace renamed a provision in quotes and missed the same word bare.
+- A command was tested only for the invocations that should fail, so it rejected every real one
+  and the suite stayed green.
+- A test asserted on a helper rather than on the code that calls it, three separate times. **A
+  test that cannot fail when the behaviour is deleted is not defending the behaviour.**
+
+## Rules
+
+- **Read the host's log before theorising.** A declaration that was sent and not applied says so
+  there and nowhere else.
+- **A failing test is kept, not skipped.** It is the reproduction.
+- **Never rotate during an adoption.** Rotation is a separate act, afterwards, deliberately.
+- **A data directory is never removed by the mesh** ([ADR 0030](../../02-DECISIONS/0030-data-outlives-the-mesh-that-declared-it.md)),
+  and that protects against the mesh only — not against a disk or a mistaken command.

00-META/process/07-feature-branches.md

@@ -0,0 +1,69 @@
+# Playbook 07 — Feature branches across repos
+
+**Trigger.** Work that changes code — in one code repo or in several at once (`mesh-sdk`,
+`mesh-controller`, `mesh-catalog`, `mesh-host`, `mesh-lab`, and `hq` when a decision rides along).
+
+**Who runs it.** Anyone who writes code, engineers and agents alike. Agents follow it exactly —
+it is the guard against the failure it was written for.
+
+## The failure it prevents
+
+A feature was worked as a branch-and-MR per *unit of thought* — one per decision, one per
+stacked increment — and each MR was treated as finished when it was *opened*, not when it was
+*merged*. Across repos the same feature took a different branch name in each. The MRs piled up
+unmerged: one session left **sixteen** stacked intermediate MRs that had to be consolidated and
+closed by hand. An MR is a review checkpoint, not a scratchpad.
+
+## The rule
+
+One feature is **one branch name**, **one worktree per repo**, **one MR per repo**, opened
+**once, at the end**.
+
+1. **Name the feature once.** `feat/<slug>`. The *same* branch name in every repo the feature
+   touches — never a different name per repo, never a fresh branch per increment within the
+   feature.
+2. **Isolate each repo.** One git worktree per touched repo under `.work/<slug>/<repo>`, branched
+   off `main`:
+   ```
+   git worktree add .work/<slug>/<repo> -b feat/<slug> origin/main
+   ```
+   Parallel features never collide, and no shared checkout is edited.
+
+   **Add the untouched siblings the lab reads.** The lab beds find the other repositories by
+   sibling path from the lab checkout (`../mesh-tools/module.json`, `../mesh-sdk`, …), the way
+   the main layout has them. A `.work/<slug>/` directory holding only the touched repos fails a
+   bed at once with `no manifest for mesh-tools at …/.work/<slug>/mesh-tools/module.json`, after
+   genesis has already passed. Give the directory those repos as **detached worktrees on `main`**
+   — never symlinks:
+   ```
+   git worktree add --detach .work/<slug>/mesh-tools main
+   ```
+3. **Commit as you go — locally.** Increments land on the one branch. Nothing is pushed and no
+   MR is opened mid-feature.
+4. **Finish, then publish.** When the whole feature is done — every repo, tests green — push
+   every branch and open **one MR per touched repo**, together.
+5. **Merge promptly, once approved.** Every merge into `main` is notified and approved
+   ([ADR 0023](../../02-DECISIONS/0023-approval-is-the-checkpoint.md)); once it is, merge —
+   do not leave it sitting. The branch is deleted on merge.
+6. **Leave nothing behind.** After the MRs merge, no `feat/<slug>` branch and no `.work/<slug>`
+   worktree survive.
+
+## What this is not
+
+- **Not a licence to batch unbounded work.** A feature is a *bounded* unit; if it sprawls for
+  days, end-of-feature bloat merely replaces per-increment bloat. Split it into features, each
+  its own branch and MR.
+- **Not a second trunk.** Every repo branches off `main`. There is no longer an `initialization`
+  trunk.
+
+## How it is checked
+
+The end state is visible, and its absence is the smell:
+
+- After a feature merges, `git branch -r | grep feat/<slug>` and `git worktree list` return
+  nothing for it. A surviving branch or worktree means step 6 was skipped.
+- A bed run from `.work/<slug>/mesh-lab` that fails naming a `.work/<slug>/<repo>/…` path it
+  cannot find is a missing sibling worktree, not a mesh fault.
+- More than one open MR in a repo that share no feature name, or a stack of MRs none of which is
+  merged, is the failure this playbook exists to prevent — stop and consolidate before opening
+  more.

00-META/repos.md

@@ -1,6 +1,6 @@
 ---
 status: canonical
-updated: 2026-08-23
+updated: 2026-09-24
 ---
 
 # The Novox repositories
@@ -17,21 +17,24 @@ and a forge address is an operational detail (see [`README`](../README.md)).
 | `hal` | The monorepo — the node runtime, the module catalogue, the delivery machinery, and the bootstrap scripts. Every core module lives here. |
 | `hq` | This repository, under the company organisation — mission, research, design, decisions, issue diagnosis. Company-scoped ([ADR 0019](../02-DECISIONS/0019-how-this-repository-works.md)); the mesh is its first product. The source of truth for *why*. Carries no implementation. |
 | *(one per application)* | Every standalone application, site or side-project gets its own repository, with `module.yml` at the root. Registered with the mesh as a build source; built and deployed by the same pipeline as anything in the monorepo. |
+| `migration` | **Private.** The record of one installation replacing the predecessor mesh with this one: the runbook, a dated log of every step and what it cost, the per-service data procedures, the readiness checks, and the scripts. Private because it is the opposite of this repository in every way that matters — it names machines, addresses, ports and paths, because a procedure that cannot be followed is not one. Where hq asks *what did we decide and why*, that repository answers *what happened on the machines, in what order, and what to do next*. Its `HANDOFF.md` is where somebody picking the work up starts. |
 
 ## What the mesh becomes
 
 [ADR 0019](../02-DECISIONS/0019-how-this-repository-works.md) records the repositories the
-monorepo decomposes into. **`mesh-lab`, `mesh-host` and `mesh-control` exist so far** — the lab is built first
+monorepo decomposes into. **`mesh-host`, `mesh-controller`, `mesh-catalog`, `mesh-lab`, `mesh-sdk` and `mesh-tools` exist
-([ADR 0016](../02-DECISIONS/0016-the-lab.md)); the rest are the
+so far** — the lab was built first ([ADR 0016](../02-DECISIONS/0016-the-lab.md)). The tiered
-target, not the present.
+decomposition below is the planned shape; the repositories built to date do not map onto it
+one-for-one — `mesh-catalog`, `mesh-sdk` and `mesh-tools` exist where the table names
+`mesh-foundation` and `mesh-surfaces`, and reconciling the two is itself still ahead.
 
 | Repository | Tier | Holds |
 |---|---|---|
 | `mesh-host` | 0 | **exists.** The node host — one statically linked binary, requiring nothing present ([ADR 0005](../02-DECISIONS/0005-the-node-host.md)) |
-| `mesh-substrate` | 1 | the four pinned services, as declarations |
+| `mesh-foundation` | 1 | the four pinned services, as declarations |
-| `mesh-control` | 2 | **exists.** The control plane and its contexts — one of seven built ([ADR 0006](../02-DECISIONS/0006-the-substrate-and-the-control-plane.md)) |
+| `mesh-controller` | 2 | **exists.** The controller and its contexts — one of seven built ([ADR 0006](../02-DECISIONS/0006-the-substrate-and-the-control-plane.md)) |
 | `mesh-surfaces` | 3 | tools, web, cli |
-| `mesh-sdk` | — | contracts shared across tiers |
+| `mesh-sdk` | — | the stable spine modules build against — the tool-serving harness, the messaging/event framework, the contracts and core primitives. Holds nothing per-module and nothing volatile ([ADR 0039](../02-DECISIONS/0039-what-the-sdk-holds-and-refuses.md)). |
 | `mesh-lab` | — | **exists.** The lab — scenario lifecycle, networking, placement. Ships to nobody; runs on a workstation. |
 
 Tier 4's shape is open, and deliberately so: see ADR 0019 and

01-RESEARCH/003-service-supervision/00-overview.md

@@ -3,7 +3,6 @@ status: graduated
 initiated: 2026-08-22
 touches: [03-DESIGN/00-as-is/05-runtime-and-installation.md]
 became:
-  - 02-DECISIONS/0005-the-node-host.md
   - 02-DECISIONS/0005-the-node-host.md
   - 03-DESIGN/01-to-be/05-the-node-host.md
 ---

01-RESEARCH/006-mesh-from-scratch/00-overview.md

@@ -90,5 +90,5 @@ the catalogue where modules genuinely change together under one intent. The skel
 | One repository per tier, or per context? | Already open from ADR 0001 as "catalogue destination — one repository or many". The skeleton assumes per tier and does not settle it. |
 | ~~Does an unprivileged node earn a place in the inventory, or only a presence?~~ | **Answered 2026-08-25** by the operator: a node is a *managed machine inside the mesh*, not an unprivileged something — and a disconnected node is still a node, in a different situation. The question posed a class distinction; the answer is that there is none, and what varies is **state**. Recorded as [ADR 0004](../../02-DECISIONS/0004-a-node-and-how-it-joins.md). |
 | ~~Does absorbing overlay, filtering, packages, supervision and the container runtime make the host too large?~~ | **Answered 2026-08-25** — [`host-size.md`](host-size.md). Measured: the absorption is smaller than the machinery that already applies state, and eight of ten adapters already carry no dependency. The risk is not size but direction, and it is two modules wide. The claim survives with its scope corrected — the host carries one concern, *apply declared state on this machine*, of which the six are instances. Recorded as [ADR 0005](../../02-DECISIONS/0005-the-node-host.md), designed in [`05-the-node-host.md`](../../03-DESIGN/01-to-be/05-the-node-host.md). |
-| ~~Four substrate services or five?~~ | **Answered conditionally**, which is the honest form — [`07-the-substrate.md`](../../03-DESIGN/01-to-be/07-the-substrate.md). The substrate is *what the control plane consumes and cannot grant itself*. The identity provider qualifies only if the control plane delegates authentication; if it authenticates natively it is an ordinary hosted service. The count follows from a decision not yet taken, and asserting four was asserting that decision. |
+| ~~Four substrate services or five?~~ | **Answered conditionally**, which is the honest form — [`07-the-foundation.md`](../../03-DESIGN/01-to-be/07-the-foundation.md). The substrate is *what the control plane consumes and cannot grant itself*. The identity provider qualifies only if the control plane delegates authentication; if it authenticates natively it is an ordinary hosted service. The count follows from a decision not yet taken, and asserting four was asserting that decision. |
 | Does `feature` survive? | The skeleton splits it in two and argues the conflation is what makes the delivery pipeline hard to reason about. Unproven. |

01-RESEARCH/011-the-module-graph/00-overview.md

@@ -4,8 +4,8 @@ initiated: 2026-08-25
 became:
   - 02-DECISIONS/0009-modules-and-the-graph.md
   - 02-DECISIONS/0008-a-context-owns-its-store.md
-  - 03-DESIGN/01-to-be/06-the-control-plane.md
+  - 03-DESIGN/01-to-be/06-the-controller.md
-  - 03-DESIGN/01-to-be/07-the-substrate.md
+  - 03-DESIGN/01-to-be/07-the-foundation.md
 touches:
   - 02-DECISIONS/0009-modules-and-the-graph.md
   - 02-DECISIONS/0009-modules-and-the-graph.md

01-RESEARCH/012-the-minimum-viable-node/00-overview.md

@@ -7,6 +7,9 @@ touches:
   - 03-DESIGN/01-to-be/05-the-node-host.md
   - 03-DESIGN/00-as-is/05-runtime-and-installation.md
   - 01-RESEARCH/011-the-module-graph/00-overview.md
+  - 02-DECISIONS/0078-the-store-and-broker-are-modules.md
+  - 02-DECISIONS/0088-the-foundation-filters-before-anything-listens.md
+  - 02-DECISIONS/0100-a-node-in-use-is-adopted-before-it-is-converged.md
 ---
 
 # 012 — The minimum viable node, and adopting what is already there
@@ -180,6 +183,16 @@ same makes a node where something the mesh needed never happened indistinguishab
 where a log level differed. Whether a failed line still lets adoption complete is therefore
 reopened by adding severity, and is not decided here.
 
+## Adoption as a mode, for the migration
+
+*2026-09-22.* The migration from the predecessor mesh gave the middle state a length. A machine
+running the predecessor is **adopted** when the mesh comes up on it — the predecessor's control
+stopped, the machine's firewall and files kept in force, the mesh opening what it needs through
+them — and stays adopted while its modules migrate one at a time, until the operator **converges**
+it. Measured on the control-node, and weighed against the alternatives, in
+[*migrating a node that is in use*](migrating-a-node-in-use.md); decided in
+[ADR 0100](../../02-DECISIONS/0100-a-node-in-use-is-adopted-before-it-is-converged.md).
+
 ## Open questions
 
 | Question | Why it is open |
@@ -191,6 +204,7 @@ reopened by adding severity, and is not decided here.
 | Can a module say which of its settings are load-bearing? | The question that dissolves the conflict rule rather than choosing a side. A setting the module *requires* cannot be kept from the machine without producing something installed and broken; a setting it merely *prefers* should always yield. Until a module can say which is which, adoption is defaulting in the dark. Belongs with the graph. |
 | Does a `failed` line still let adoption complete? | *Flags inform, they do not block* was decided about conflicts, where the mesh chose and the machine works. A failure is *we could not*, which is different in kind — and treating them alike hides the worse one behind the commoner one. |
 | How is a flagged conflict reconciled, and by whom? | The briefing hands it to a session. What that session is empowered to change, and whether the resolution is recorded so the next adoption does not re-raise it, is undecided. |
+| ~~How long is a machine adopted?~~ | **Decided** ([ADR 0100](../../02-DECISIONS/0100-a-node-in-use-is-adopted-before-it-is-converged.md)) — for as long as it is being migrated: a mode per node, recorded, ended by an explicit and previewed flip. |
 | Where does the kept original live, and for how long? | Whether it is recorded in the node's state so adoption is visibly reversible, and whether it is returned when the mesh stops managing the thing. |
 | What shape is a briefing? | Structured enough to be acted on, prose enough to be read. It is the first thing a session on a new node sees, which makes it an interface rather than a log. |
 | Does owning a package mean owning its version? | Owning configuration and owning the package are different scopes. The second means the mesh decides which version is installed, and that decision then has to survive the machine's own package manager updating it. |

01-RESEARCH/012-the-minimum-viable-node/migrating-a-node-in-use.md

@@ -0,0 +1,154 @@
+# Migrating a node that is in use — adoption as a mode, not a moment
+
+*2026-09-22. Measured on the machine that will be the control-node, which is running the
+predecessor mesh today. Nothing below names it; the counts are its own.*
+
+## The question
+
+The mesh replaces a predecessor mesh that is running, on the same machines, with the services
+people use. The control-node is decided: it is the machine that already carries the predecessor's
+broker and build pipeline. So the question is not *where* the mesh starts but **how a machine
+running the predecessor becomes a node of the mesh without its services noticing** — and then how
+each of the other machines follows.
+
+## The operator's proposal
+
+Proposed by the operator, 2026-09-22, and the shape this document tests:
+
+1. **Stop the predecessor's control on a machine** — its daemons that write configuration: the
+   network and firewall configuration above all, which decide what is reachable and what is
+   blocked. Its services keep running; only the control over their configuration stops.
+2. **Bring the mesh up on that machine in adoption mode.** It takes custody of those files and
+   keeps what it finds in force.
+3. **Migrate the modules one at a time**, data preserved, per the cutover procedure.
+4. **Move to the next machine and repeat** — adopted first, keeping its local configuration, then
+   migrated.
+5. **When every machine is migrated, flip adoption mode**, and the mesh takes full control of the
+   configuration it has been holding.
+
+This is the research above made concrete. It keeps the conflict rule already decided here — *on
+conflict, what is on the machine stays* — and gives the middle state, *adopted*, a length: not a
+one-time import before generating starts, but a mode that lasts for as long as the machine is
+being migrated, ended by an explicit act.
+
+## What the machine actually looks like
+
+Measured, read-only:
+
+| | |
+|---|---|
+| Containers running | 60, all the predecessor's services and their stores |
+| The predecessor's control | user-level daemons, separate from the services; none of the 20 running system services is the predecessor's control |
+| Firewall | the predecessor's, active: 54 incoming rules and 52 forwarding rules, each served port allowed explicitly — how a default-deny firewall reads |
+| Files the predecessor's configuration sync writes | 12, of which 3 are system files (an ssh server drop-in, the package manager's configuration, one service's configuration); the rest are the operator's shell and agent files |
+| Per-service configuration | environment and composition files per service, written by the predecessor's service tooling |
+
+**Stopping the predecessor's control stops nothing that serves.** The services are containers and
+system units that run without it; what stops is the rewriting of their configuration. Nothing
+changes on the machine until something else writes.
+
+## What collides, measured rather than assumed
+
+An earlier note assumed the mesh's foundation could not stand beside the predecessor because both
+want the store's and the broker's standard ports. **The measurement says otherwise.** The
+predecessor publishes its own store on a non-standard port and its broker on another; the standard
+ports the foundation binds for its store and its bus are free.
+
+What does collide:
+
+| The mesh wants | Held by | When |
+|---|---|---|
+| the registry's port | the predecessor's registry | at genesis — the foundation raises a registry |
+| the broker's management port, on loopback | the predecessor's broker | at genesis |
+| the private network's port | the predecessor's own tunnel | at genesis on a control-node that is the private network's hub; otherwise when the node is placed on it |
+| the web ports | the predecessor's reverse proxy | when the route proxy is assigned |
+| the resolver's port | a resolver the predecessor runs | when a resolver module is assigned |
+
+On the control-node, which is the private network's hub, three of these fall at genesis and cannot
+be deferred; two only when a particular module is taken, the moment its predecessor stops anyway.
+Today the foundation's ports are **fixed**: written in the installer's bundle and in the
+catalogue's manifests, so a collision is found when a container fails to bind, not before — and a
+port changed at genesis would be changed back when the foundation is adopted as modules, since the
+applier recreates a container whose declared spec differs. The private network's address range
+must also stay clear of the range the predecessor's tunnel uses; on the machines measured they are
+distinct.
+
+## What would break if the mesh came up as it is today
+
+**The firewall.** Genesis loads a base ruleset — drop anything undeclared — in its own table
+([ADR 0088](../../02-DECISIONS/0088-the-foundation-filters-before-anything-listens.md)). The
+predecessor's firewall is a different table. The kernel runs every base chain registered at the
+same hook, in priority order: an accept ends only its own chain and the packet goes on to the next,
+and a drop in any is final — whether the other firewall's chains are nftables or legacy iptables.
+So the
+mesh's base ruleset would drop everything the predecessor's firewall allows and the mesh has not
+declared — every web, mail and database port in the table above — the moment genesis ran.
+
+**The files.** The host writes a declared file whatever it finds at the path, reporting it as
+updated. A file the predecessor left — the ssh server drop-in, the resolver's configuration — is
+replaced the first time a module declaring that path is assigned, before that module's service
+has moved.
+
+**The container names.** The applier keys a container on its name. A catalogue module whose
+container carries the same name as the predecessor's service it replaces takes that container over
+the moment it is assigned: today, *assigning a module is migrating it*, never a preparation.
+
+**The published ports.** The foundation publishes its ports on every interface, and a published
+container port reaches the container through the forwarded path, not the incoming one. A firewall
+that filters only incoming traffic never sees it. The base ruleset is what keeps the store
+unreachable from outside today — and it is the thing that cannot be loaded on this machine.
+
+## The pipeline freezes while the control-node migrates
+
+The predecessor's build pipeline and its coordinator run on the control-node. Stopping its control
+there stops the predecessor's updates for **every** machine it manages, until the migration is
+done. Their services keep running; they receive nothing new. That is the price of the proposal and
+it is worth stating, not a reason against it: the migration is the period in which the predecessor
+is being replaced, and it does not need to keep changing.
+
+## What adoption mode has to mean
+
+For the proposal to hold, *adopted* must be a **state the mesh records per node**, not an
+intention, and each thing the mesh would otherwise take must say what it does in that state:
+
+- **What is found is kept until its module is taken.** *Found* is precise: present at a declared
+  path or name with no record in the host's store. A found file or container is held, its original
+  recorded, until the operator **takes** the module on that node — the cutover, done when the
+  module's data has moved. Assigning prepares; taking migrates. Without the distinction the rule
+  never fires: the host only ever sees what assigned modules declare.
+- **The firewall found on the machine stays in force.** The mesh loads no table on an adopted node
+  that drops by default or holds an accept. What it needs open it declares as openings the host converges
+  *through the found firewall*, on the incoming and the forwarded path, marked as the mesh's and
+  re-checked on every reconcile so a reload or reboot does not lose them. An accept in a table of
+  its own would not help: the found firewall's drop would still be final. What a table of its own
+  *can* do is refuse, and a refusal is final too — so the mesh guards the store and the broker's
+  management port from everyone but the private network and the machine itself, in a table that
+  only refuses, ahead of the container runtime's redirect — which the found firewall does not do
+  and cannot undo. The bus, the registry and the hub's port stay open to
+  anywhere: a node enrols before it has a private-network address.
+- **The foundation's ports are the node's to give** — set at genesis, checked free, and kept as
+  that node's settings, read everywhere they are used, so adopting the foundation as modules does
+  not move them back.
+- **The flip is per node and previewed** from what is actually reachable — listening sockets and
+  published ports, not the found firewall's allow list, which does not see what a container runtime
+  forwards.
+
+## Options weighed
+
+| Option | Verdict |
+|---|---|
+| Cut the machine over in one go: stop the predecessor's store, broker, registry and proxy, raise the foundation in their place | Rejected. Every predecessor service goes down until it has migrated, and the predecessor's other machines lose their broker. The rollback is restarting the predecessor, which is a recovery, not a step. |
+| A separate machine as the control-node | Rejected. Contradicts the decision that the control-node is the machine that already carries the predecessor's broker and pipeline. |
+| Converge on joining, as the mesh does today | Rejected. The base ruleset closes every predecessor port at genesis, and found files are replaced before their services move. |
+| Make only the foundation's ports configurable and otherwise converge | Rejected as insufficient. It solves the bind collisions and none of the firewall or file ones. |
+| Treat assigning a module as migrating it | Rejected on review. The rule that keeps found files would never fire — the host only sees what assigned modules declare — and an assignment on an adopted node would be an outage rather than a preparation. |
+| **Adoption as a mode, per node, ended by an explicit flip** | The proposal. It is the conflict rule already decided here, given a duration. |
+
+## What this leaves open
+
+Answered by the decision record this feeds — [ADR 0100](../../02-DECISIONS/0100-a-node-in-use-is-adopted-before-it-is-converged.md)
+— only for the migration's needs. The general questions above stay open: whether a module can say
+which settings are load-bearing, what shape a briefing takes, how a flagged conflict is reconciled.
+One is newly sharp: the mesh opening ports through a firewall it did not install needs to speak that
+firewall. There is one kind on the machines measured; a machine with another is not covered until
+someone writes for it.

01-RESEARCH/013-the-forge-and-the-registries/00-overview.md

@@ -0,0 +1,54 @@
+---
+status: active
+initiated: 2026-09-23
+touches:
+  - 02-DECISIONS/0075-two-stores-and-which-provides-what.md
+  - 02-DECISIONS/0079-the-foundation-seats-are-named-after-their-servers.md
+  - 02-DECISIONS/0071-genesis-builds-from-a-mesh-that-already-exists.md
+  - 02-DECISIONS/0082-the-registry-is-reached-by-name-and-trusted-by-the-overlay.md
+  - 04-ISSUES/085-the-packages-port-given-at-genesis-is-not-a-setting/00-report.md
+  - 04-ISSUES/090-the-forge-module-does-not-take-over-the-forge-genesis-raised/00-report.md
+---
+
+# 013 — The forge and the registries: what a seat is for, and what it is not
+
+**The question, asked during the first migration:** the mesh runs a forge that serves git and
+packages, an image registry, and a bootstrap forge genesis raises before any module exists. Should
+there be a mesh-scoped **git seat**, the way there is one for the store and the broker?
+
+**The short answer: no, and the question points at a different hole.** A seat answers *is there
+exactly one of you*. Every symptom around the forge is about **an address and a credential nobody
+resolves**, which a seat does not answer. The mechanism that would answer it — a provision, with
+`serves` and a grant — already exists, is already used for the image store, and is already
+declared for packages. It simply has no consumer: the one module that needs it carries a literal
+instead.
+
+See [the survey](the-survey.md) for what the code does today, with evidence.
+
+## What was found
+
+- **A seat does exactly two things**: it refuses a second claimant at resolution, and in one place
+  it answers *where is the broker*. It is a resolution-time predicate; no host ever hears of it.
+- **Four mesh-scoped seats exist**, all named after a singular server the mesh runs **for its own
+  working** — controller, store, broker, catalogue. A forge is an application the world made, not
+  part of that set.
+- **Nothing in the mesh requires git.** There is no provision for it, and the forge's git service —
+  over https and ssh — is declared nowhere. Only its npm half is a provision.
+- **`package-registry` is a fully built provision with zero consumers.** The builder, its only
+  real consumer, bypasses it with a file naming the module, the address `127.0.0.1` and the port.
+- **The builder is told where source lives, per build, by a person.** The clone URL is an opaque
+  string; each module remembers its own; no credential is ever attached; nothing polls a forge.
+- **A seat would forbid something the mesh should allow**: a second forge — one for the mesh's own
+  source, one for something else — which ADR 0075 already argues for against a node-scoped claim.
+
+## What this leaves open
+
+- Should the forge's git service become a **provision** (`source-forge`), so the builder resolves
+  the address and a credential the way it already resolves the image store? That is what would let
+  a forge move, be renamed, or be replaced without editing every module's stored URL, and it is the
+  unanswered half of [issue 085](../../04-ISSUES/085-the-packages-port-given-at-genesis-is-not-a-setting/00-report.md).
+- The obstacle is known and written down: such a requirement **may go unanswered during genesis**,
+  and the mesh has no optional requirement. Deciding that is the real work.
+- Should the mesh learn when a source moves? It records the fact and never discovers it: nothing
+  polls, and there is no receiver for a forge's push. `build --behind` answers a question only a
+  person can currently make true.

01-RESEARCH/013-the-forge-and-the-registries/the-survey.md

@@ -0,0 +1,130 @@
+# The survey — seats, the forge, the registries, as the code has them
+
+*2026-09-23. Read from the code and the records, not from memory. Every claim here was checked
+against a file.*
+
+## What a seat is
+
+The glossary calls a seat *a named position at a scope with a capacity*, and a claim *a module
+taking a spot on it*. **Capacity does not exist in the code.** There is no capacity field and no
+bench; every claim is exclusive. The "shared seat" of the design is the word `provides` wearing
+that name.
+
+Claiming does two things and no others:
+
+- **It refuses a second claimant when a node's modules are resolved** — within one node at any
+  scope, and against every other node for mesh and site scope.
+- **It answers one question, once:** the controller finds where the broker is by looking for the
+  module claiming the broker's seat.
+
+No host is ever told about a seat. It is a resolution-time predicate.
+
+Two mechanical asymmetries matter. A mesh-scoped seat's exclusivity is defended only by nodes
+that *resolve*: a node whose modules fail to resolve contributes nothing, so a broken node does
+not hold its seat against a second claimant. And a site-scoped seat does nothing at all when
+either node has no site.
+
+## The seats that exist
+
+Twelve claims, twelve names. Four are mesh-scoped: the controller, the store, the broker, the
+catalogue. All four are things the mesh runs **for its own working**, and three of them are raised
+by genesis and adopted in place — which is the argument of the record that named them.
+
+The node-scoped ones split in two: a **role on this machine** (the build machine, the packet
+filter, the intrusion prevention, the private network) and **a scarce machine resource** (the
+resolver's configuration file, the DNS port). The second kind is a claim for the reason a port is:
+there is one of it on the machine.
+
+Two oddities worth stating:
+
+- The image store claims its seat at **node** scope while providing its service at **mesh** scope.
+  The record that discussed this left open whether it should hold that claim at all.
+- **Nothing claims the two public ports.** The proxy binds them and claims nothing, which is the
+  hole the route handover fell into.
+
+## The forge, and what it serves
+
+The forge serves three things: git over https, git over ssh on its own port, and an npm registry.
+**Only the npm half is in the mesh's vocabulary** — the forge provides `package-registry` and says
+where it answers. Git is served and declared nowhere: no provision, no `serves`, nothing that can
+require it. The only trace is the public label in its route contribution, which the mesh is
+explicitly not meant to interpret.
+
+The image store is a separate module and a separate provision, plain HTTP, trusted because it is
+reachable only over the private network.
+
+A second module also provides `package-registry`. Two providers of one provision is not a refusal
+but an ambiguity, and the mesh already has the command that settles it: pin one.
+
+## Where the builder's three addresses come from
+
+| What it needs | How it finds it |
+|---|---|
+| the broker | a sealed secret of its own |
+| the image store | **a real provision binding**, resolved by the mesh |
+| the package registry | **a file in its own manifest**, naming the module, `127.0.0.1` and the port |
+| the source repository | nothing at all — a person types a clone URL per build |
+
+The installer says this plainly in a comment: *the one binding nothing resolves: the builder dials
+the forge by a number it carries.* Genesis can override the port of that literal, and nothing
+else: not the forge's identity, not its address.
+
+`package-registry` therefore has **zero consumers**. The builder does not require it. The
+provision is fully built — two providers, grants, a served description including the registry's
+path — and nothing asks for it.
+
+## How a module's source is found
+
+It is not found; it is supplied. The clone URL arrives as an argument, travels to the build machine
+and reaches `git clone` unexamined. No credential is attached, so a private forge works only if the
+build machine's own git configuration already authenticates. Each module remembers the URL it was
+built from, so there are as many forge addresses as modules, each frozen at whatever was typed.
+
+At genesis there are four independent source URLs, one per flag, unrelated to each other.
+
+If the forge moves or is renamed: every stored URL goes stale independently, each failing at its
+own clone, and there is no command to re-point them. Genesis keeps working, because it takes URLs
+as flags — which is the existing record's position: the forge is reached *by a name outside the
+mesh*, and failover is repointing that name.
+
+## What genesis's forge is
+
+A bare container run before any module exists, because the toolchain resolves the mesh's own
+packages by version from an npm registry. It holds **no seat, no manifest, no provision**. The
+mesh is told one thing about it — a port — and only when that port is not the default. **On an
+ordinary genesis the mesh never learns the bootstrap forge exists.**
+
+Its successor module differs from it in five ways: container name, network, data directory, the
+address it is told to call itself, and — contradicting a comment that says they are pinned
+identically — **the image digest**. So assigning the module raises a second forge beside the first.
+
+A seat cannot fix that. A seat is checked between manifests; the bootstrap forge has no manifest,
+so no seat can see it. The check that is actually wanted — same name, same data, same network — is
+between an installer constant and a manifest.
+
+## Would a git seat be coherent?
+
+In form, yes: add the claim and a second forge is refused. In effect it buys one refusal nobody
+has hit, and forbids an arrangement the mesh should allow — a second forge for something other
+than the mesh's own source.
+
+Measured against the four mesh seats, it does not fit: those are singular servers the mesh runs
+for itself, and a forge is an application. Nothing would *use* it, because the only seat consumer
+answers "where is the broker", and a forge address is not one fact: a module's source is a
+repository **and a path and a ref**.
+
+## What the code already supports and nobody uses
+
+- **`package-registry` with no consumers.** The builder's literal file is a hand-rolled copy of
+  the binding the mesh would have written for it.
+- **Pinning a provider** already expresses *this* forge, of several, without forbidding the second.
+- **Grants** on the forge are the credential mechanism that would hand the builder an account.
+  Today the installer creates that account directly, because nothing requires the provision and so
+  the grant has no consumer to write for.
+
+## The reading
+
+The question *should there be a git seat* is the wrong shape for the problem under it. Every
+symptom — the literal address, the port that would not follow a node, the takeover that is not a
+takeover — is about **an address and a credential nobody resolves**. A seat resolves neither. The
+provision mechanism does, and is already built.

01-RESEARCH/014-the-bus-on-nats/00-overview.md

@@ -0,0 +1,108 @@
+---
+status: graduated
+became: 02-DECISIONS/0106-the-bus-is-nats.md
+initiated: 2026-09-23
+touches:
+  - 02-DECISIONS/0002-nodes-communicate-over-a-broker.md
+  - 02-DECISIONS/0033-the-substrate-is-a-store-and-a-broker.md
+  - 02-DECISIONS/0039-what-the-sdk-holds-and-refuses.md
+  - 02-DECISIONS/0041-events-are-a-relationship.md
+  - 02-DECISIONS/0042-the-shape-of-an-event-on-the-wire.md
+  - 02-DECISIONS/0043-a-module-broker-account-is-scoped-by-emits-and-consumes.md
+  - 02-DECISIONS/0078-the-store-and-broker-are-modules.md
+  - 02-DECISIONS/0083-one-push-leaves-the-mesh-consistent.md
+---
+
+# 014 — The bus on NATS: replace the broker, and when
+
+**The question, asked mid-migration:** the operator wants the mesh's bus — today an AMQP broker —
+replaced by NATS, with everything the broker does today. Do we finish the migration on AMQP and
+move to NATS after, or go to NATS directly?
+
+**The short answer: decide NATS now, build it in the lab in parallel, and cut the mesh's bus over
+in one rehearsed rollout — after the migration's core is done and never underneath it.** "First
+everything on AMQP" is already the state and costs nothing more: the predecessor's broker was
+merged into the mesh's tonight, and every module converted from here on targets the sdk's broker
+contract, which names no protocol. The only code that speaks AMQP is the mesh's own, in three
+places, and it is swapped once.
+
+## What was measured
+
+**Where AMQP is spoken** (non-test files): the controller's `internal/link` (5 files, ~2.4k lines
+with tests) and the builder's main; the host's `internal/link` (~1.5k); the tool runtime's
+`broker-amqp.ts` (~600 with its main). **The sdk speaks none** — its `Broker` is `request`,
+`handle`, `publish`, `subscribe`, `close` ([ADR 0039](../../02-DECISIONS/0039-what-the-sdk-holds-and-refuses.md)
+put the client in the runtime for exactly this). **Every module's tools and events go through the
+sdk**; one module (`amqp-ping`, a probe) talks AMQP on purpose.
+
+**What the bus carries:** a control queue (`control`, `.upgrades`, `.catchup`), one queue per node
+(`node.<name>`), `builds`, the enrolment/report/alive/built flows, a topic exchange of events
+(`mesh.events.<module>.<event>`, dead-lettered to `mesh.events.dead`), an RPC exchange (`mesh.rpc`)
+and per-tool service queues (`serve.<module>.<tool>`), plus an MQTT exchange.
+
+**What it relies on:** TLS for the bus a node enrols over; prefetch with reject/nack so the control
+queue **holds messages unacknowledged while the store restarts** and retries
+([ADR 0083](../../02-DECISIONS/0083-one-push-leaves-the-mesh-consistent.md)); publisher confirms and
+mandatory routing; a dead-letter exchange; **per-module accounts scoped by `emits`/`consumes`**
+([ADR 0043](../../02-DECISIONS/0043-a-module-broker-account-is-scoped-by-emits-and-consumes.md)),
+minted through the broker's management API; the broker as a module holding the `mesh-broker` seat
+([ADR 0078](../../02-DECISIONS/0078-the-store-and-broker-are-modules.md)).
+
+**The predecessor's world**, now on the same broker: ~50 AMQP connections per machine from four
+machines, 595 queues, four vhosts, six users. All of it AMQP, all of it retiring module by module.
+NATS speaks no AMQP: that world cannot move; it does not need to.
+
+## How NATS answers each of those
+
+| The mesh relies on | NATS | Note |
+|---|---|---|
+| topic routing keys | subjects with wildcards | same shape (`mesh.events.>`) |
+| RPC and tool invocation over an exchange + reply queue | request/reply, native | simpler than today |
+| competing consumers | queue groups | same |
+| durable control queue, hold-unacked-and-retry, catch-up | JetStream: streams, durable consumers, ack/nak with delay, replay | the 0083 guarantee moves to JetStream; core NATS alone is at-most-once and would not do |
+| dead-letter | max-deliver + advisories, or a stream fed from them | different mechanism, same effect |
+| TLS bus | TLS | same |
+| accounts scoped by emits/consumes, vhosts | accounts (isolation) with users and per-subject publish/subscribe permissions | stronger than today; the controller writes an auth config the host declares and the server reloads, instead of calling a management API |
+| management API | `nats` CLI and an HTTP monitoring endpoint | no vhost concept — accounts instead |
+| MQTT | built in | same |
+| a module seat `mesh-broker` | unchanged — the seat is the server, the module changes | ADR 0079 |
+| multi-node | clusters and leaf nodes | not needed now; a leaf per node is a later question |
+
+Nothing the mesh needs is missing. The differences are in the shape of durability (JetStream must
+be declared, streams and consumers are objects) and of accounts (configuration, not API calls).
+
+## The cost, honestly
+
+The bus is the mesh's nervous system. Moving it means: a new `nats` module in the catalogue taking
+the `mesh-broker` seat; the controller's and the host's link packages rewritten; the tool runtime's
+client swapped behind the unchanged contract; enrolment, reports, builds and the guard's ports
+re-derived; the lab beds that prove the bus (store window, enrolment, upgrades) re-run on the new
+one; and eight decisions amended or superseded. Weeks, not days — and none of it can be done
+halfway on a live mesh: the controller, every host and every tool runtime move together.
+
+## The sequencing question, answered
+
+1. **NATS first, directly.** Stalls the migration for the length of the build; the mesh's bus
+   changes under a half-migrated node; the predecessor's clients still need AMQP, so a second
+   broker runs anyway, plus a bridge for whatever crosses. Rejected.
+2. **Finish on AMQP, NATS after.** Wastes nothing — no module written from here on speaks AMQP —
+   but leaves the decision unmade while modules are written, and the runtimes idle. Adequate.
+3. **Decide NATS now; build it in the lab in parallel; cut the mesh's bus over in one rehearsed
+   rollout after the core is migrated.** The predecessor's clients never notice: their broker is
+   the one the mesh adopted, kept as a compatibility module with an end date — the day the last
+   AMQP client is gone. **Recommended.**
+
+## What this leaves open
+
+- The **decision itself**, as a record: the bus is NATS; the AMQP broker becomes the predecessor's
+  compatibility broker and retires with the last AMQP client. Written when the operator says so.
+- **JetStream's shape for the control plane**: one stream per concern (control, nodes, builds,
+  events) or one with subjects; retention; what the store window guarantee looks like as ack-wait
+  and nak-delay. Measured in the lab, not designed on paper.
+- **Accounts as configuration**: the controller writes users and permissions into a file the host
+  declares, reloaded on change — which is the [ADR 0102](../../04-ISSUES/102-an-address-recorded-at-genesis-or-build-does-not-follow-the-nodes-ports/00-report.md)
+  discipline applied from the start — or the JWT/operator model. The first is simpler and matches
+  how the mesh already writes everything.
+- **The MCP bridge** the operator asked for the same day is written against the sdk's contract, so
+  it moves with the bus and is not written twice.
+- Whether a **leaf node per machine** replaces the hub-and-spoke bus later — out of scope here.

01-RESEARCH/015-the-object-store-after-minio/00-overview.md

@@ -0,0 +1,142 @@
+---
+status: active
+initiated: 2026-09-24
+touches:
+  - 02-DECISIONS/0028-the-substrate-supplies-the-control-plane-and-nothing-else.md
+  - 02-DECISIONS/0033-the-substrate-is-a-store-and-a-broker.md
+  - 02-DECISIONS/0048-a-provider-creates-the-credential-the-mesh-minted.md
+  - 02-DECISIONS/0049-a-consumers-identity-fits-the-tightest-backend.md
+  - 02-DECISIONS/0078-the-store-and-broker-are-modules.md
+  - 02-DECISIONS/0084-which-provider-serves-a-consumer.md
+  - 03-DESIGN/00-as-is/03-provisioning.md
+  - 03-DESIGN/01-to-be/07-the-foundation.md
+  - 04-ISSUES/113-the-object-stores-images-were-withdrawn-upstream/00-report.md
+---
+
+# 015 — The object store after MinIO: which S3 implementation, and how the data moves
+
+**The question.** The mesh's object store is MinIO. Its community edition is archived upstream,
+its server and client images have been deleted from every public registry, and the pinned release
+is four and a half years old and will never be patched
+([issue 113](../../04-ISSUES/113-the-object-stores-images-were-withdrawn-upstream/00-report.md)).
+Which S3-compatible implementation replaces it, and what is the migration track for the data and
+the provisioning model that sit on top of it?
+
+**Why now, and why not sooner.** Nothing is on fire: nodes that already hold the images keep
+running, and issue 113 establishes that the deploy path tolerates an unfetchable-but-present
+image by design. The forcing function is not an outage but a one-way door — **no node that does
+not already hold the images can ever provision the module again**, so the mesh's ability to stand
+a node up from its declarations is already broken for this module, and silently.
+
+**The direction is not a departure from the design; it is the design.** The foundation document
+already states the commitment:
+
+> The dependency is on the **protocol**, not the product: AMQP for the bus, S3 for the object
+> store, the OCI protocol for the registry. That is what keeps the naming safe rather than a
+> commitment that cannot be revisited.
+
+The object store is also **not** a foundation service — ADR 0028 removed it, and it is an
+ordinary module required through the module graph by whatever wants one. (The "exception that is
+not a swap" in that passage is the relational store, whose provisioning model borrows PostgreSQL's
+own meaning of databases, roles and schemas. The object store carries no such coupling: a bucket
+is a bucket.) So this effort is an instantiation of an existing principle, not a redesign — which
+is the cheapest kind of decision to make and the strongest kind to cite.
+
+## What the replacement has to carry, measured
+
+Taken from the module's manifest, its composition, its tool surface, and a search for its
+consumers across the catalogue — not from assumption.
+
+| Requirement | Evidence in the module today |
+|---|---|
+| S3 API | The protocol every consumer speaks; already the design's stated dependency. |
+| ~~OIDC login against the mesh's identity provider~~ | **Struck 2026-09-24. Not a requirement, and it never worked.** Six variables are wired and an entrypoint blocks on the provider, which reads as a live feature. The module's own hook comment records the end state as *"policy claim missing"* — a failing login. See [01](01-candidate-comparison.md). |
+| **Per-application access keys, each scoped to a bucket** | The real requirement. A "user" of the store is normally an application; the mesh already mints a credential per provisioned bucket. |
+| **One live consumer using it as opaque primary storage** | A file-sync application, since early 2023: objects named by internal id, metadata in its own database. Highest-risk consumer — a live copy drifts, and its bucket name must be preserved. |
+| Erasure-coded multi-node topology | Four server nodes with two data directories each, behind a load balancer. |
+| A single-node form | Declared as a flavour, for development and small nodes. |
+| Buckets as a typed provision | The module declares a provision type of `bucket` on a named network; the mesh mints the credential and the provider creates it (ADRs 0048, 0084). |
+| A tool surface | Bucket create/list/delete, object list/info/delete, presigned URL, and provisioning. |
+| A console | Published on its own subdomain through the reverse proxy, with an unlimited request-body middleware for uploads. |
+
+**Consumers, counted:** one application module, one capture module that takes a private bucket per
+node, one workflow module's tools, and the delivery/rescue internals of the shared library. The
+surface is small — the cost is concentrated in the provisioning handler, the tool handlers and the
+OIDC story, not spread across the catalogue.
+
+## Candidates
+
+**Four candidates, not three.** The comparison was briefly narrowed to SeaweedFS on the strength of
+console single sign-on; that axis turned out not to be a requirement, and the incumbent's own
+maintained fork had been omitted altogether. Both errors, and why they happened, are recorded in
+[01 — the candidates measured](01-candidate-comparison.md), which carries the evidence and the
+requirement-by-requirement detail.
+
+In short, and only in short:
+
+- **The maintained fork of the incumbent** — the community edition was archived and its images
+  deleted, but a fork publishes, tracks CVEs, and preserves the on-disk format, S3 API and
+  environment surface. Costs **an image reference** where every other option costs a data
+  migration, two rewrites and a maintenance window. Does not end the dependence on an abandoned
+  codebase; buys time to choose deliberately.
+- **Garage** — its permission model *is* the requirement (per access key, per bucket), its admin
+  API is the closest match to how the mesh provisions, and the highest-risk consumer is
+  first-party documented against it. Remaining cost: no object versioning, no server-side
+  encryption or object locking, partial lifecycle — **unmeasured against the ten buckets, and the
+  one thing that could still disqualify it**.
+- **SeaweedFS** — longest field record and erasure coding. Its console sign-on is a paid feature,
+  which is now beside the point. What weighs against it is narrower: its S3 surface is a gateway
+  translating onto its own file-system API, with no first-party support for the opaque consumer.
+- **RustFS** — closest in shape to the incumbent, so the least porting. But it reached general
+  availability eight days before this was written, and carries an open defect in the credential
+  path. Two earlier claims about it are corrected in 01: it is **not** a drop-in that retains
+  existing data.
+- **Ceph RGW** — remains rejected as disproportionate where the object store is an ordinary
+  module rather than a platform.
+
+**This is now two decisions, not one:** whether to repoint to the fork or migrate, and — if
+migrating — to which. Repointing does not foreclose migrating, which is the argument for taking it
+first. On the corrected requirement the migration ranking is Garage, then SeaweedFS, and not yet
+RustFS. Two measurements gate any graduation: **which S3 endpoints the consumers actually call**
+(Garage cannot be ranked fairly until counted), and **whether the fork can read the incumbent's
+on-disk format in place** — tested on a copy, because the migration between them is one-way. Both
+are in [01](01-candidate-comparison.md#what-is-still-unmeasured).
+
+## The migration track, in outline
+
+Data movement is the easy half, and deliberately reversible.
+
+1. **Stand the replacement up beside the incumbent**, on its own ports, its own provision type and
+   **its own data directory**. Nothing removed. The data directory matters: reusing one the
+   incumbent already holds would put a fresh single-drive store on top of a live erasure set.
+2. **Copy bucket by bucket with a neutral tool.** `rclone` rather than the incumbent's own client
+   — the client has been withdrawn upstream too, so building the migration on it would inherit
+   the same dependency this effort exists to remove.
+3. **Verify per bucket** — object counts and checksums, not a transfer exit code.
+4. **Repoint consumers through the connection the module already publishes.** Consumers read an
+   API URL from the module's declared connections rather than addressing the store directly, so
+   the cutover surface is that value plus the provisioning and tool handlers.
+5. **Freeze writes, final incremental sync, flip**, and keep the incumbent read-only as the
+   rollback until confidence is earned. For the opaque consumer this is **not optional and not
+   instant**: it stores objects by internal id with metadata in its own database, so a copy taken
+   while it runs will drift. It needs a maintenance window for the final sync, and the window is
+   proportional to 82,496 objects rather than to 230 GiB.
+6. **Retire**, and only then remove the module.
+
+The genuinely new work is not the copy. It is the **provisioning handler** and the **tool
+handlers**, both written against the incumbent's admin API. *The OIDC wiring was previously listed
+here and is struck: it is not a requirement and it never worked.*
+
+## Open questions
+
+- ~~How much of the OIDC requirement survives, and in which build?~~ **Answered, and it was the
+  wrong question.** The console requirement does not exist, and the login it referred to never
+  worked. What replaced it: which S3 endpoints consumers actually call, and whether the fork reads
+  the incumbent's format in place.
+- Does the mesh's bucket provision translate to the candidate's identity model without weakening
+  what ADR 0049 says about a consumer's identity fitting the tightest backend?
+- Should this effort also answer issue 113's general question — mirroring third-party images into
+  the mesh's own registry — or is that a separate decision? Replacing one withdrawn product with
+  another unmirrored upstream leaves the same one-way door in place, just further from the hinge.
+- Is the four-node erasure-coded topology still warranted, or was it inherited? Worth re-asking
+  while the product is being chosen, rather than reproducing a shape by default.

01-RESEARCH/015-the-object-store-after-minio/01-candidate-comparison.md

@@ -0,0 +1,176 @@
+# 015 / 01 — The candidates measured
+
+*Rewritten 2026-09-24. An earlier version of this document ranked the candidates on whether they
+preserved single-sign-on to the object store's **console**. That was the wrong axis — it is not a
+requirement — and a fourth candidate was missing entirely. Both errors are recorded at the end,
+because how a comparison came to be ranked on the wrong thing is worth more than the ranking was.*
+
+## The requirement, corrected
+
+Taken from the operator and from the running system, not from the module's shape.
+
+**A "user" of the object store is normally an application.** The requirement is therefore
+**per-application access keys, each scoped to its own bucket** — not per-human single sign-on. The
+mesh already works this way: it mints a credential for every provisioned bucket, and the consumer
+reads an endpoint from the module's declared connection rather than addressing the store directly.
+
+**The console is not a requirement.** It was the axis the previous version ranked on, and it should
+not have been.
+
+**The identity-provider login never worked.** The predecessor's module wires six OIDC variables and
+blocks startup until the provider answers, which reads like a working feature. It is not: the
+module's own hook comment records the end state as *"policy claim missing"* — a **failing** login,
+written up as progress because it proved the provider had registered. The identity provider emits no
+such claim, nothing in the module creates the mapper, and the configured scope alone would not carry
+a custom one. Two days of logs show no genuine login attempts, only internet scanners failing on an
+STS API version. **Nothing should be carried forward on the assumption this works**, and no
+candidate should be credited or penalised for matching it.
+
+**One consumer is live, opaque, and holds real user files.** A file-sync application has used the
+store as its **primary storage** since early 2023: objects named by an internal id, with all
+metadata in its own database. Three consequences — a copy taken while it runs will drift, its bucket
+name must be preserved or its database references break, and it is the highest-risk consumer of the
+lot.
+
+## What is actually stored, measured
+
+| | |
+|---|---|
+| Logical | **230 GiB, 82,496 objects, 10 buckets** |
+| Raw on disk | **468 GiB** — eight drive directories at 59 GiB each |
+| Implied scheme | 468 ÷ 230 = **2.03×**, confirming erasure coding at half parity |
+| Headroom | ~1.3 TiB free on the filesystem holding it |
+
+**All eight "drives" are directories on one filesystem on one machine.** The erasure coding is
+therefore not buying independent-drive redundancy; the real failure domain is the array underneath,
+which has its own. This single fact decides more of the comparison than any product feature: a
+scheme's redundancy model is close to irrelevant here, and what remains is its storage overhead.
+
+At 230 GiB with 1.3 TiB free, **storage overhead is not a deciding cost either.** Replication at
+three copies would run ~690 GiB against the present 468 GiB — about **+222 GiB**, comfortably
+absorbed. Erasure coding at a wider stripe would *save* roughly 146 GiB. Both are rounding errors
+against the headroom, and neither should decide this.
+
+## The candidates
+
+Four, not three. The previous version omitted the first.
+
+### The maintained fork of the incumbent
+
+The community edition was archived upstream and its images deleted
+([issue 113](../../04-ISSUES/113-the-object-stores-images-were-withdrawn-upstream/00-report.md)),
+but **a fork is maintained and publishing** — `pgsty/minio`, from the Pigsty project. It restores
+the console stripped from the community
+build, rebuilt image and package distribution, tracks CVEs, and states that it preserves the on-disk
+format, the S3 API and the environment-variable surface. Verified by pulling it: it reports a
+current release, permissive-to-copyleft licensing unchanged from upstream, and identifies itself as
+a community fork. Adoption is real — the server image has been pulled three quarters of a million
+times.
+
+**Why it reorders the comparison.** Every other candidate costs a data migration, a provisioning
+handler rewritten against a different admin API, a tool surface ported, and a maintenance window for
+the opaque consumer. The fork costs **an image reference**. It also closes the issue's one-way door:
+a node holding nothing can provision the module again, and patches resume.
+
+**What it does not do** is end the dependence on a codebase its original authors abandoned. It is
+maintenance mode, largely one project's effort, with no new features intended. It buys time to
+choose deliberately rather than under pressure — which is worth a great deal, and is not the same as
+a decision.
+
+### Garage
+
+**The best fit for how the mesh provisions.** Its permission model is *per access key, per bucket,
+read/write/owner* — which is the requirement above stated verbatim rather than approximated. Its
+admin API is a first-class REST surface with tokens scopeable to exactly the two operations a bucket
+provision performs. The opaque consumer is **first-party documented** against it, for primary
+storage, including client-side encryption support.
+
+Its previously-recorded penalties mostly dissolve under the corrected requirement: it has no console
+and no identity-provider integration, neither of which is wanted; and it replaces AWS-style ACLs and
+bucket policies with its own per-key-per-bucket model, which is the thing being asked for.
+
+**What genuinely remains.** It replicates rather than erasure-codes — immaterial at this volume and
+on a single array, as above. It does **not implement the full span of S3 endpoints**: object
+versioning is absent, object locking and server-side encryption endpoints are absent, and lifecycle
+is partial. **Whether any of the ten buckets depends on those is unmeasured, and it is the one thing
+that could still disqualify it.**
+
+### SeaweedFS
+
+Longest field record of the group, permissive licence, erasure coding, and identity-provider
+integration on the S3 API through token exchange. **Its console sign-on is a paid feature** — the
+admin UI itself is open, its identity integration is not. That finding is what falsified the
+previous version's narrowing, and it is now largely beside the point, since the console is not a
+requirement.
+
+What weighs against it here is narrower and more specific: its S3 surface is a **gateway
+translating onto its own file-system API**, with acknowledged divergence from AWS behaviour at the
+edges, and there is no first-party documentation for the opaque consumer. For a store already
+holding real user files in an opaque layout, first-party support is worth more than a feature list.
+It also carries more moving parts than a single-machine deployment needs.
+
+### RustFS
+
+Closest in shape to the incumbent — a similar admin API and client compatibility, so the existing
+handlers would port with least effort — under a permissive licence, with erasure coding and a
+console that does integrate an identity provider.
+
+Two things were recorded about it earlier that were **wrong, and are corrected here**: it is *not* a
+binary-level drop-in that retains existing data (API compatibility and on-disk compatibility are
+separate paths, and the on-disk one is preview-scoped with documented encryption limits), and it
+therefore offers no shortcut around the migration. It also carries **an open defect in the exact
+area the mesh depends on** — an access key created by an identity-provider user reported denied on
+all S3 operations.
+
+Decisively for now: **it reached general availability eight days before this was written.** For a
+component holding 230 GiB of real user files, field record is a feature, and it does not have one.
+
+### Ceph RGW
+
+Remains rejected, for the reason already recorded: disproportionate where the object store is an
+ordinary module rather than a platform.
+
+## Where this leaves it
+
+**The decision is no longer "which product replaces the incumbent".** It is two decisions, and they
+can be taken in either order but should not be confused:
+
+1. **Repoint to the maintained fork, or migrate now?** Repointing is an image reference and it
+   closes the issue. Migrating now costs a data copy, two rewrites and a maintenance window, and
+   buys independence from an abandoned codebase sooner.
+2. **If migrating, which?** On the corrected requirement the ranking is **Garage first** — its
+   permission model *is* the requirement, its provisioning API is the closest match, and the
+   highest-risk consumer is first-party supported. SeaweedFS second, on field record, with a
+   translation-layer caveat that matters more here than its feature list. RustFS not yet, on age.
+
+Taking (1) does not foreclose (2), and that asymmetry is the argument for taking (1) first.
+
+## What is still unmeasured
+
+1. **Whether any of the ten buckets needs object versioning, server-side encryption or lifecycle.**
+   This gates Garage specifically and nothing else here answers it.
+2. **Whether the fork's release can actually read the incumbent's on-disk format in place.** The
+   format is claimed compatible across a multi-year gap; the migration between them is one-way, so
+   this is tested on a copy or not at all.
+3. **Whether the opaque consumer's maintenance window is acceptable**, and how long it actually is
+   at 82,496 objects.
+4. **Whether the eight-drive erasure-coded shape is warranted at all.** The evidence above says it
+   is not buying what it appears to: eight directories, one array, one machine. It looks inherited.
+
+**Nothing graduates to a decision before 1 and 2.**
+
+## Two errors in the previous version of this document
+
+Recorded because the shape of both survives anonymisation and neither is unique to this effort.
+
+**It ranked on a requirement that did not exist.** Console single sign-on was treated as the axis
+because the module's configuration showed it wired up, and a wired-up configuration was read as a
+used feature. It was neither used nor working. *A configured feature is not an observed one*, and
+the evidence needed was the operator's answer and the logs — both cheap, neither consulted before
+the ranking was written.
+
+**It omitted the incumbent's own fork.** The whole effort began because an upstream withdrew its
+images; whether anyone had continued that upstream was the first question to ask and it was not
+asked. The candidate list was assembled from a search for *alternatives*, which by construction
+returns things that are not the incumbent. *When a dependency dies, "who took it over" precedes
+"what replaces it".*

01-RESEARCH/016-how-a-credential-can-be-rotated/00-overview.md

@@ -0,0 +1,57 @@
+---
+status: graduated
+became:
+  - 02-DECISIONS/0114-a-shared-credential-rotates-over-two-credentials.md
+  - 03-DESIGN/01-to-be/27-a-module-requires-the-mesh-resolves.md
+initiated: 2026-09-26
+touches:
+  - 02-DECISIONS/0113-the-vault-makes-every-secret.md
+  - 02-DECISIONS/0049-a-consumers-identity-fits-the-tightest-backend.md
+  - 02-DECISIONS/0048-a-provider-creates-the-credential-the-mesh-minted.md
+  - 03-DESIGN/01-to-be/13-credentials-and-their-rotation.md
+  - 03-DESIGN/01-to-be/27-a-module-requires-the-mesh-resolves.md
+  - 04-ISSUES/103-a-container-is-not-recreated-when-a-file-it-reads-changes/00-report.md
+---
+
+# 016 — How a credential can be rotated
+
+**What.** Which rotation mechanisms the mesh's providers can actually support, measured against
+their code rather than assumed. Every provider in the catalogue was read, found by listing every definition that provides something: how it names what it
+makes for a consumer, what its remove destroys, whether it re-applies a password, whether its
+backend can hold two secrets for one login or two logins on one resource, and how its own
+administrative credential is set. The consumer side was read too: when a module reads a secret, and
+what makes it read a new one.
+
+**Why.** [ADR 0113](../../02-DECISIONS/0113-the-vault-makes-every-secret.md), as first drafted,
+chose *overlap*: add a second login beside the first, move every reader, then remove the old one,
+"through the adapter's existing create and remove", with "no consumer changes". A review showed that
+claim false. In most providers the consumer's data is named after its login, and remove drops the data
+with the login. Overlap as written would have deleted every consumer's database on its first
+rotation. The mechanism has to be chosen on what the providers do.
+
+**What it touches.** Rotation in 0113 and [to-be 27](../../03-DESIGN/01-to-be/27-a-module-requires-the-mesh-resolves.md),
+which [ADR 0114](../../02-DECISIONS/0114-a-shared-credential-rotates-over-two-credentials.md) decided on
+these findings. The identity budget in
+[ADR 0049](../../02-DECISIONS/0049-a-consumers-identity-fits-the-tightest-backend.md), if a consumer
+gets two logins. The rotation already implemented, which [to-be 13](../../03-DESIGN/01-to-be/13-credentials-and-their-rotation.md)
+describes.
+
+**Documents.**
+
+- [01 — The providers](01-the-providers.md): the survey, one row per provider, and what it shows.
+- [02 — The readers](02-the-readers.md): how a secret reaches a running process, and what already
+  recreates it.
+- [03 — The options](03-the-options.md): each rotation mechanism against those facts, and a
+  recommendation.
+
+**Finding, in one paragraph.** All nine credential providers already re-apply a consumer's password
+in place on every create, and the controller's `rotate` command relies on that. It is a working
+rotation with a stated window. Eight of the nine name the consumer's resource after its login, and five
+destroy the consumer's data when they remove the login. The harness, keyed by login, would do the same
+on any change of login. Only one backend holds two passwords on one login, and two more hold several
+tokens. Eight backends can grant two logins the same rights over one resource; the ninth can give one
+login a second token. So every provider can hold **two credentials** over one resource, but only after
+each adapter separates *the consumer's resource* from *the credential that reaches it*. In postgres
+that also means the resource belongs to a role no login owns. Administrative credentials are a
+different case. They have one party and a fixed name, and five backends take them only at first
+initialisation, so changing one needs the old and the new value at once.

01-RESEARCH/016-how-a-credential-can-be-rotated/01-the-providers.md

@@ -0,0 +1,103 @@
+# 01 — The providers
+
+Read from the catalogue's main branch: each provider's provisioner adapter (`create`, `remove`),
+the client functions they call, and each definition's own credentials. The providers were found by
+listing every definition that provides something and has a provisioner, not from memory. A first pass
+of this survey worked from memory and missed one, mailu.
+
+**The provisioner harness** in `mesh-sdk` calls `create` for a consumer when its contribution appears
+or changes (its login, password or values), and after the provisioner restarts. It calls `remove` for
+a login it applied earlier in the same process that is no longer contributed. Its record of what was
+applied is kept in memory and keyed by login. Two things follow:
+
+- a consumer whose derived login changes is removed under the old login and created under the new one,
+  in one pass;
+- a contribution that disappears while the provisioner is down is never removed, and is left behind.
+
+## The credential providers
+
+`login` is the consumer's derived identity, which the adapter receives as `as`
+([ADR 0049](../../02-DECISIONS/0049-a-consumers-identity-fits-the-tightest-backend.md)).
+
+| provider | the consumer's resource is named | remove destroys | create re-applies the password | two secrets on one login | two logins on one resource |
+|---|---|---|---|---|---|
+| postgres | a database named `login`, owned by the role `login` | the database and the role | yes, `ALTER ROLE … PASSWORD` when the role exists | no: a role has one password | yes, but only through a role that cannot log in owning the database, with each login working as it. Otherwise whatever one login creates is its own, and dropping that login means handing its objects over first. Not done today |
+| mssql | a database named `login`, with the login mapped into it | the database and the login | yes, `ALTER LOGIN … WITH PASSWORD` | no: a login has one password | yes, two logins mapped to users in `db_owner`. A user owning a schema cannot be dropped, and a login with an open session cannot. Not done today |
+| mongodb | a database named `login`, with a user holding `dbOwner` | the database and the user | yes, `updateUser` with the new password | no: a user has one credential | yes, two users with `dbOwner` on one database. Not done today |
+| redis | the key prefix `login:` on an ACL user named `login` | the user, **not** its keys | yes: `ACL SETUSER … reset … >password` replaces all of them | **yes**: an ACL user holds several passwords, added with `>` and removed with `<`. Today's `reset` discards all but the new one | yes, two users on one key prefix, once the prefix is not the login |
+| minio | a bucket derived from `login`, and a service account whose access key is `login` | the access key; the bucket **only if empty**. A bucket holding objects is left, and the failure logged | yes, by removing the access key and adding it again, which leaves a moment with no key | no, but an access key *is* the login: a second key is a second login | yes, two service accounts with one bucket policy. The access key is capped at 20 characters |
+| lavinmq | a virtual host named `login`, and a user named `login` with permissions on it | the virtual host, with any queued messages, and the user | yes, the user is written again with the password | no: a user has one password | yes, permissions for two users on one virtual host |
+| mosquitto | a client named `login`, with a role named for it on the topic prefix `login/#` | the client and its role | yes, the password is set when the client exists | no: a client has one password | yes, two clients holding one role, once the prefix is not the login. The MQTT client identifier is chosen by the consumer, not tied to the login; a duplicate one takes the older session over |
+| mailu | a mailbox `login@domain`, unless the consumer contributes its own account name | the mailbox with its mail, for a login-named one; a contributed name is left for an operator | yes, the password is set when the user exists | no for the password; a user can hold several authentication tokens, per the backend's documentation | **no**: a mail user *is* its mailbox |
+| gitea (npm) | a user named `login` on a team of an organisation that owns every package | the user; **packages survive**, because the organisation owns them | yes, the user's password is set on every run | no for the password; a user can hold several access tokens | yes, trivially: a second member of the same team |
+
+## The other providers
+
+| provider | answers with | credential |
+|---|---|---|
+| umami | a website, found by its public name | none. The site id it makes has no way back to the consumer today |
+| cloudflare-dns | a public name derived from `login` | none handed to the consumer; its own API token is an operator value |
+| showcase | a route | none |
+| mesh-vault | custody: it records and withdraws sealed values in a ledger | it holds secrets; it makes none today |
+
+verdaccio provides the npm registry too, and has no provisioner.
+
+## Each provider's own administrative credential
+
+| provider | identity | how the backend takes it |
+|---|---|---|
+| postgres | a fixed superuser | from a file **only at first initialisation** |
+| mssql | `sa` | from the environment at first setup. The image documents no file form, and the definition records that as a declared exception |
+| mongodb | a fixed `root` | from a file **only at first initialisation**, when the data directory is empty |
+| mosquitto | a fixed admin client | seeded into the broker's dynamic-security file **once**; the seeding step skips when the file exists |
+| lavinmq | a fixed admin name | per its own bootstrap code, **only on a first boot** with an empty data directory. No resource in the definition runs that bootstrap; what sets it on a running mesh is outside the catalogue |
+| redis | the default user | from `requirepass` in a configuration the mesh renders, read when the server starts |
+| minio | a fixed root user | from a file, read when the server starts |
+
+**In five of seven, a new administrative value takes effect only through a command run with the old
+one.** The credential file is mounted directly into both the server and the provisioner. So replacing
+it recreates the provisioner, which then holds only the new value while the backend still expects the
+old one, and the provisioner is locked out. That is worse than changing nothing.
+
+Every provider module also has its own bus account, an own secret, read at start.
+
+## What the tables show
+
+1. **Every credential provider already rotates in place.** All nine re-apply the password on the
+   same login each time `create` runs. The controller's `rotate` command relies on that: it replaces
+   the credential in the inventory and sends both ends in one push. Its own comments state the window,
+   between the provider applying and the consumer restarting, in which the consumer cannot
+   authenticate.
+2. **Eight of nine name the consumer's resource after its login.** Only gitea separates them,
+   because an organisation owns the packages. A second login therefore has no resource of its own to
+   reach, and cannot share the first one's without the adapter granting it.
+3. **Five of nine destroy the consumer's data when they remove the login**: postgres, mssql and
+   mongodb drop the database, lavinmq drops the virtual host with its queued messages, and mailu
+   deletes the mailbox with its mail. minio drops only an empty bucket, and redis leaves the keys. In
+   those five, *retire a login* and *delete the consumer's data* are one call. With the harness keyed
+   by login, a changed login triggers it too.
+4. **One backend holds two passwords on one login** (redis). Two hold several tokens beside one
+   password (gitea and mailu). A rotation built on two secrets per login would work for three
+   providers out of nine.
+5. **Eight of nine can give two logins the same rights over one resource.** Group roles in postgres,
+   database roles in mssql and mongodb, permissions in lavinmq, a shared role in mosquitto, a shared
+   policy in minio, a shared key prefix in redis, a shared team in gitea. mailu cannot, because its
+   user is its mailbox, but it can give one user a second token. So every provider can hold **two
+   credentials** over one resource, though not every one as two logins. No adapter does either today.
+6. **Ownership is a trap in two backends.** In postgres whatever a login creates is that login's, so a
+   second login cannot alter the first one's tables, and the first cannot be dropped while it owns
+   them. The one-step way out deletes them. In mssql, a login cannot be dropped with a session open,
+   nor its user while it owns a schema.
+7. **The administrative credentials have one party and a fixed name**, and five backends take them
+   only at first initialisation. The provisioner needs the old and the new value at once to change
+   them. Today nothing can give it both.
+8. **A consumer's identity is already the resource's name.** The login is derived from the
+   assignment, which is a module on a node, so the current login and "the consumer" are the same
+   string today. A second login would need a new name. The resource can keep the one it has.
+
+## Seen on the way
+
+The redis configuration names no ACL file, so a consumer's ACL user exists only in memory. A restart
+of the redis server erases every consumer's user. The provisioner does not create them again until it
+restarts itself, because its in-memory record says they are done. That is not a rotation finding, but
+it is a live fault, and it is recorded here so it is not lost.

01-RESEARCH/016-how-a-credential-can-be-rotated/02-the-readers.md

@@ -0,0 +1,46 @@
+# 02 — The readers
+
+How a secret reaches a running process, and what makes the process take a new one.
+
+## No module watches a secret
+
+A search of every module's code in the catalogue found no file watching of any kind, and no
+re-reading of a secret while running. **Every reader reads a secret when it starts.** There is no
+consumer that takes a new value live, so every rotation that changes what a consumer presents ends in
+the consumer restarting.
+
+## The host already recreates what read a changed file
+
+The node host records, for every long-running container, the digest of each file it read when it
+was created: its env-files, and every file bind-mounted into it directly. When a digest changes, the
+host recreates the container, even though its spec is otherwise unchanged. This is the fix for
+[issue 103](../../04-ISSUES/103-a-container-is-not-recreated-when-a-file-it-reads-changes/00-report.md).
+It is on the host's main branch, while the issue is still recorded as located, not fixed.
+
+Two cases are deliberately left out and need `restart-on` in the definition:
+
+- a file read out of a **mounted directory**, because the host cannot know whether the service reads
+  it once or watches it (a route proxy re-reads its routes live; a provisioner polls what it receives);
+- a **process** rather than a container.
+
+## What the catalogue does with it
+
+22 definitions declare a secret they receive. In 16 of them it reaches the service through a
+rendered file, usually an env-file. That case the host already covers. 14 declare `restart-on` for
+something. Whether each of the 22 is fully covered depends on how its secret travels: through an
+env-file or a direct mount, which the host covers, or through a directory or into a process, which
+needs `restart-on`. **That was not classified module by module.** It is the check to run before a
+rotation mechanism relies on it.
+
+The count covers only the `secrets` field. The 49 modules with their own bus account, and 54 with any
+own secret, are readers too, and their bus accounts are rotated like any credential two parties hold.
+Their files are mounted directly, which the host covers, but the classification has to name them.
+
+## What this means for rotation
+
+- The *read at start* half of 0113's recipient model is already true, and mostly already handled by
+  the host. The restart is derived from the files a container reads, not declared per secret.
+- Any mechanism, in place or overlapping, ends with the reader being recreated. What differs is
+  whether the credential it held until then still works.
+- For a single-party secret, a module's own, the reader is also the only holder. There is nobody to
+  overlap with, and delivering the new file recreates the reader.

01-RESEARCH/016-how-a-credential-can-be-rotated/03-the-options.md

@@ -0,0 +1,79 @@
+# 03 — The options
+
+Three mechanisms, weighed against [01](01-the-providers.md) and [02](02-the-readers.md).
+
+## A. In place, as today
+
+The vault makes a new value. Every applier re-applies it on the same login, which all nine
+providers already do. Every reader is recreated by the host.
+
+- **Works with:** every provider, unchanged. It is what `rotate` does now.
+- **Costs:** a window per consumer, from the provider applying to the consumer being recreated. They
+  are on different machines, and nothing orders them. A reader whose machine is unreachable from the
+  mesh but still reaches its provider stays locked out until the mesh reaches it again.
+- **Admin credentials:** the natural form. The provider module is the only party, and it has to
+  apply the new value with the old one anyway (finding 6).
+
+## B. Two secrets on one login
+
+The applier adds the new password beside the old one, readers move, and the old one is removed.
+
+- **Works with:** redis natively, and gitea and mailu through tokens. **Not** with the other six, whose
+  backends hold one password per login (finding 4).
+- **Verdict:** not a mechanism, a special case. Using it where it exists and something else
+  elsewhere is the "this way or that way" the design is trying to remove.
+
+## C. Two credentials per consumer, over one resource
+
+The consumer has two credentials and uses one at a time. The applier ensures the other with the new
+value and gives it the same rights over the consumer's resource. Readers move to it, and then the old
+credential is retired, which removes the credential only, never the resource. **What a credential is,
+is the adapter's**: a second login for eight providers (finding 5), a second token on the same login
+for mailu. The mesh sees one mechanism.
+
+- **Works with:** every provider, **after** each adapter changes:
+  - the resource is named after the consumer, not the login. Today the two are the same string (finding
+    8), so existing resources keep their names, and the current login stays one of the two;
+  - the resource is owned by the resource, not by a login. In postgres that is a role no one logs in
+    as, which each login works as, and ownership of an existing database moves to it once (finding 6);
+  - both credentials get the same rights, over data and structure;
+  - *retire a credential* and *remove the consumer* become two operations. Today they are one call, and
+    in five providers that call destroys data (finding 3). The harness must key by consumer, so that a
+    changed login is not a removal. This is the whole of the danger, and it has to be split, whatever
+    else is chosen.
+- **Costs:**
+  - every credential adapter changes;
+  - the harness learns the alternation and a confirmation per step, and rotation state has to live
+    somewhere that survives a restart, which the harness's memory does not;
+  - the second login's name must fit the tightest backend. That is 20 characters for a minio access
+    key ([ADR 0049](../../02-DECISIONS/0049-a-consumers-identity-fits-the-tightest-backend.md)), and
+    a suffix spends part of it;
+  - retiring an mssql login has to end its sessions first.
+- **Gains:** no window. A reader that cannot be reached keeps a working credential until it can.
+- **Does not apply to** single-party secrets: admin credentials and a module's own secrets. There is
+  no second party to overlap with.
+
+## Independent of the choice
+
+- **Split remove, and key the harness by consumer.** Retiring a credential, or a login changing, must
+  never be able to destroy a consumer's data. That holds under A too, because A's remove is the same
+  call.
+- **Admin credentials are applied by their own provider**, using the old value, with the new one staged
+  beside it. Five backends take the value only at first initialisation. Replacing the file first locks
+  the provisioner out (finding 7).
+- **Classify the readers** ([02](02-the-readers.md)), bus-account readers included, before relying on
+  derived restarts.
+
+## Recommendation
+
+- **Two-party credentials, consumer credentials and bus accounts: C**, because it is the only
+  mechanism every provider supports, and it closes the window instead of shortening it. Its prerequisite, separating the resource from the login and
+  retiring a login from removing a consumer, is worth doing on its own, because it removes a
+  data-loss path that exists today.
+- **Single-party secrets (admin credentials, a module's own): A, staged.** In place, applied by the
+  provider that holds them, with the new value beside the old until it has taken.
+- **Until the adapters are changed, A stays** as `rotate` implements it, with its window stated. It is
+  not replaced by a mechanism the providers cannot yet carry.
+
+This is two mechanisms, split by a property of the secret rather than by provider: whether it has one
+party or two. Every provider is treated the same way for the same kind of secret.

01-RESEARCH/017-a-mesh-that-heals-itself/00-overview.md

@@ -0,0 +1,47 @@
+---
+status: active
+initiated: 2026-09-26
+touches:
+  - 00-META/mission.md
+  - 02-DECISIONS/0106-the-bus-is-nats.md
+  - 02-DECISIONS/0010-delivery.md
+  - 02-DECISIONS/0083-one-push-leaves-the-mesh-consistent.md
+  - 03-DESIGN/01-to-be/06-the-controller.md
+  - 03-DESIGN/01-to-be/09-the-node-lifecycle.md
+  - 03-DESIGN/00-as-is/09-interfaces-and-observability.md
+---
+
+# 017 — A mesh that heals itself
+
+**What.** The behaviour the operator wants: a mesh that runs itself. It notices what is wrong,
+repairs what it can, and hands what it cannot repair to someone who can, with the reason. This effort
+writes that wish down as intended behaviour, designed for the bus the mesh is moving to
+([ADR 0106](../../02-DECISIONS/0106-the-bus-is-nats.md): NATS). It also records what can be done
+pragmatically before that move.
+
+**Why.** The mission is *a mesh that controls itself* ([mission](../../00-META/mission.md)). The
+mesh can tell whether it is up. It cannot tell whether it is right. The as-is page on observability says so
+([as-is 09](../../03-DESIGN/00-as-is/09-interfaces-and-observability.md)). To-be 06 names an
+`observability` context in the controller and leaves its store undecided. Nothing routes a condition
+the mesh cannot fix to anyone. The cost is measurable: **46 of the 116 issue reports in this
+repository describe a failure that was silent.** A mesh that heals itself is, first, a mesh that stops
+failing silently.
+
+**What it touches.** The controller's observability context, the node lifecycle's liveness, delivery
+([ADR 0010](../../02-DECISIONS/0010-delivery.md), [ADR 0083](../../02-DECISIONS/0083-one-push-leaves-the-mesh-consistent.md)),
+the provisioner harness, and rotation, which is proposed alongside to-be 27 as ADR 0114.
+
+**Documents.**
+
+- [01 — The intended behaviour](01-the-intended-behaviour.md): the wish, as principles and as how
+  the mesh behaves once the bus is NATS.
+- [02 — Now, pragmatically](02-now-pragmatically.md): what is done before NATS, why it does not
+  build anything the move would throw away, and what has been done already.
+
+**Next.** Two measurements this effort owes before it can graduate:
+
+1. **Every loop in the mesh**: what it converges, and whether it compares against observed state or
+   against its own memory. Issue 120 found the provisioner harness trusting memory. The same pattern is
+   expected elsewhere.
+2. **The 46 silent failures, classified**: a missing observation, a loop trusting memory, or a missing
+   escalation. That shows which mechanism removes the most of them.

01-RESEARCH/017-a-mesh-that-heals-itself/01-the-intended-behaviour.md

@@ -0,0 +1,85 @@
+# 01 — The intended behaviour
+
+The operator's wish, written as behaviour: what a person or an agent sees the mesh doing. This is a
+target to design toward, not a design. Every part of it is to be decided through a record before it
+is built.
+
+## Principles
+
+**1. Every loop compares what should be with what is, never with what it did.** Desired state is the
+mesh's: assignments, requirements, seats. Observed state is read from the thing itself: the container,
+the backend, the node. A loop that compares against its own memory of what it applied is blind to
+anything that changed behind its back. That is issue 120, and it is the pattern this whole effort is
+written against.
+
+**2. Healing is the ordinary path run again, never a second path.** Repairing a lost login is
+provisioning it. Repairing a dead container is converging the node. Repairing a stale declaration is
+delivering it. A repair that needs its own code is a second way of doing something, which is exactly
+what the mesh is removing everywhere else.
+
+**3. A repair never destroys.** Healing may recreate, re-provision, re-deliver and restart. It may
+never delete a consumer's data, retire a credential someone still uses, or pick a winner between two
+contradictory states. Where the only repair is destructive, it is escalated.
+
+**4. Nothing fails silently.** Every condition the mesh cannot repair within its budget becomes
+visible. It is named, it says since when, why, and who can resolve it. It is visible until it is
+resolved, and resolved by observation, not by someone clicking it away.
+
+**5. What the mesh cannot fix goes to an agent.** Per the mission, an agent may be human or not. A
+condition that needs judgement is handed to one, as work, with what the mesh knows. It is not handed
+over as a notification that someone may or may not read.
+
+**6. Correctness, not only liveness.** A running process that authenticates with a dead credential,
+serves an old version, or routes nowhere is not healthy. What a provision's contract promises is what
+is checked: the credential authenticates, the route answers, the version is the declared one.
+
+## The loop, everywhere
+
+Every part of the mesh that owns something runs the same loop:
+
+1. **know** what should be true: from assignments, requirements and seats;
+2. **observe** what is true: from the thing itself, on its own cadence;
+3. **repair** the difference by running the ordinary path again, within a budget of attempts and
+   time;
+4. **raise** a *condition* when the budget is spent or the only repair is destructive;
+5. **clear** the condition when observation shows it resolved.
+
+A **condition** is a durable fact about something the mesh owns, such as a node, an assignment, a
+provision, a seat or a rotation: what is wrong, since when, the evidence, what was tried, and who can
+resolve it. Conditions are the one thing a person or an agent looks at to know whether the mesh is
+right. `status` is the list of open conditions. When it is empty, the mesh is right, not just up.
+
+## On NATS
+
+[ADR 0106](../../02-DECISIONS/0106-the-bus-is-nats.md) moves the bus to NATS, and NATS makes most of
+this cheaper, because observation becomes something every component publishes rather than something
+a central process polls.
+
+| the wish needs | on NATS |
+|---|---|
+| every component says it is alive | a heartbeat on a subject per node and assignment; silence past its interval is a condition, and nobody polls |
+| every component says what it observed | observations published on subjects (`mesh.observed.<node>.<assignment>`, for instance), consumed by whoever owns the comparison |
+| the last known state survives restarts | a JetStream key-value bucket of observed state per owner; the provisioner's "what I applied" and a rotation's step live there, not in memory |
+| conditions are durable and watchable | conditions as entries in a key-value bucket, watched by anyone who cares: a surface, an agent, the controller |
+| the bus itself is observed | the server's advisories (a consumer exceeding its deliveries, a slow consumer, a client disconnecting) and its monitoring endpoint become observations like any other |
+| a repair is retried, not lost | JetStream redelivery with delay, which is the same mechanism [ADR 0083](../../02-DECISIONS/0083-one-push-leaves-the-mesh-consistent.md)'s guarantee moves to |
+| work handed to an agent | a condition that needs judgement published as a task on a subject an agent's queue group consumes |
+
+**Who compares.** Each owner compares its own: the host for its node's containers and files, a
+provisioner for its backend, the vault for rotations, the controller for delivery and seats. The
+controller's observability context does not repair anything. It holds conditions, their history,
+and the view across the mesh. It notices what no owner can see about itself: an owner gone silent.
+
+## What stays human
+
+Some repairs need the operator's key, and the mesh says so rather than pretending otherwise:
+re-raising the vault or the broker, and recovering a node's identity. These are conditions too, with
+the procedure named, and they are the only ones that can never clear themselves.
+
+## Open
+
+- The budgets: how many attempts, over how long, per kind of repair.
+- How a condition that needs judgement reaches an agent, and how the agent's action is recorded.
+- Where the observability context stores history (to-be 06 left it open; volume argues against the
+  relational store).
+- Which correctness probe each provision's contract offers, and how often it runs.

01-RESEARCH/017-a-mesh-that-heals-itself/02-now-pragmatically.md

@@ -0,0 +1,46 @@
+# 02 — Now, pragmatically
+
+The intended behaviour lands on NATS. The bus moves after the migration's core
+([ADR 0106](../../02-DECISIONS/0106-the-bus-is-nats.md)). Until then, work toward it is chosen by one
+test:
+
+**Does it survive the move?** A change to what a loop compares against, or to what an adapter can
+tell about its backend, survives, because it is independent of the bus. A new AMQP queue for health
+reports, a poller written against the broker's management API, or a condition store built on the
+current broker does not survive, and is not built.
+
+## Done
+
+**The provisioner asks the backend, not memory** (issue 120). The SDK's harness gained an optional
+`holds` on the adapter, asked for every applied consumer every minute. A consumer the backend no
+longer holds is provisioned again. Being unable to ask is not treated as loss. The cache module
+implements it first, because its server keeps its users in memory and forgets them all on a restart.
+That was verified against a real server: a restart erases every consumer's user, and `holds` answers
+correctly for absent, present, wrong-password, disabled and deleted.
+Changes: mesh-sdk PR #7 (0.1.1) and mesh-catalog PR #84.
+
+This is principle 1 applied to one loop. It survives the move unchanged. On NATS, the harness's
+record of what it applied moves from memory into a key-value bucket, and `holds` stays as it is.
+
+## Next, in order of silent failures removed
+
+1. **`holds` for the other credential providers.** Each backend can answer whether a login exists
+   with the mesh's password without changing anything. Where a backend cannot check a password without
+   logging in, logging in is the check.
+2. **The harness's other blind spot.** A consumer that goes away while its provisioner is down is never
+   removed. The fix is the same principle in reverse: list what the backend holds, and compare it with
+   what the mesh asks for. Removal stays subject to ADR 0114's rule that it never follows from a login
+   changing.
+3. **`status` reports what owners already know.** The host knows which containers it recreated and
+   why. A rotation knows who it waits on. Delivery knows what is outstanding. Surfacing those as
+   conditions in the existing `status` needs no new transport. It is the shape the NATS condition store
+   will hold.
+4. **The loop inventory and the classification** in [00](00-overview.md). They decide what comes after
+   these three.
+
+## Not now
+
+- Heartbeats, observation subjects, key-value state, advisories: all NATS, all after the move.
+- Handing conditions to agents: designed with NATS, where a task on a subject is native.
+- Choosing the observability store: decided when there is something to store, which is after the
+  move.

02-DECISIONS/0024-model-access-is-a-provision.md

@@ -1,6 +1,6 @@
 ---
 topic: what runs on it
-status: proposed
+status: accepted
 date: 2026-08-30
 deciders: jochen
 reconstructed: false

02-DECISIONS/0025-the-design-record-is-read-not-copied.md

@@ -0,0 +1,105 @@
+---
+topic: how we work
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0019-how-this-repository-works.md
+---
+
+# 25. The design record is read where it is written, never copied to be found
+
+## Context
+
+**These documents cannot be found by searching the mesh's memory, and never could.** Checked on
+2026-08-23 and again on 2026-08-31, against both the symptom-indexed store and the structured
+archive, using a decision record's full title and a distinctive phrase from a design document: no
+result, no partial match, no stale copy.
+
+That matters because of what was promised. The objection to giving this material its own
+repository was that the mesh already has a knowledge store, and a second one repeats the mistake
+that store was created to fix. **The answer offered was indexing rather than location** — that
+these documents would be returned beside everything else in a search, so where they were authored
+became a separate question. The indexing was never built.
+
+**The claim has since stopped being load-bearing**, which is why this is a decision rather than an
+incident. [`README.md`](../README.md) names the gap in the place the claim used to sit, and
+[ADR 0019](0019-how-this-repository-works.md)'s reasoning rests on cadence, reviewers and scope —
+none of which depend on being searchable from elsewhere. What remained was an unbuilt capability
+and an open question, recorded as
+[`04-ISSUES/006`](../04-ISSUES/006-hq-is-not-indexed-into-the-knowledge-base/00-report.md).
+
+**A signpost was added on 2026-08-31 and measured.** One entry in the mesh's memory naming what
+lives here and when to come looking. A search for *design records, decisions, repository* returns
+it; a search phrased the way somebody actually asks — *why is the mesh built this way* — returns
+nothing, because the store matches terms and not meaning. **Reachable is not the same as
+surfacing**, and the measurement is what established which one a signpost buys.
+
+## Considered Options
+
+1. **A one-way sync into the mesh's memory.** A job reads this repository on a schedule and writes
+   the documents into the searchable store. It works with what exists today and needs nothing
+   built first. **Rejected**, because it creates a second copy of every document, and the failure
+   mode of a derived copy is the one this repository is least able to tolerate: *the copy that is
+   searched quietly stops matching the copy that is edited*, and the enforced one wins. A design
+   record that has silently diverged from the reasoning it claims to carry is worse than one that
+   cannot be found — the first misleads, the second merely fails.
+
+2. **Leave the signpost and close nothing.** Honest, free, and it keeps the gap visible.
+   **Rejected as an end state**, though it is what stands until the option below exists. It
+   answers only for a reader who already suspects these documents exist, which is precisely not
+   the reader the mesh's memory is designed for.
+
+3. **An agent reads this repository directly, and the search consults it.** Nothing is copied.
+   **Adopted.**
+
+## Decision
+
+**The design record is read where it is written.** Retrieval is an agent reading this repository,
+not a copy living in a second store — and a search of the mesh's memory consults that agent, so
+what it knows appears beside ordinary results rather than only when it is asked.
+
+Both halves are the decision. The first alone is merely a reader, and would leave this repository
+reachable but not surfacing — the state measured above. **The second half is what discharges the
+promise** that these documents are returned beside everything else.
+
+**There is no copy, and that is the point.** No sync, no schedule, no reconciliation, and nothing
+that can drift, because there is only ever one of each document. It is also always current,
+including for work that is not yet committed.
+
+**The direction of reading is one-way and stays that way.** The agent reads this repository and
+answers from it. Nothing flows back: this repository is public, the mesh is not, and a return path
+would be how installation-specific detail arrives into documents that must not carry it
+([`README.md`](../README.md)).
+
+## Consequences
+
+**This repository stops being a fourth knowledge system, properly.** The original objection was
+about adding a knowledge *system*. An agent with read access adds no store at all — which answers
+the objection more completely than the indexing that was promised, rather than merely as well.
+
+**ADR 0019's promise is amended, not satisfied.** It said these documents would be *indexed*. They
+will not be. They will be *read*, and the search will ask. The commitment that survives is the one
+that mattered — that a searcher finds them without already suspecting they exist — and the
+mechanism behind it is different from the one named.
+
+**It is gated on an agent that does not exist yet.** Until it does, the signpost is what stands,
+and this repository is reachable rather than surfacing. That is a known and stated gap, not a
+silent one — and the gap is now a build task with a decided shape rather than an open question.
+
+**The search must degrade honestly.** When the agent cannot be reached, a search has to say that
+this material was not consulted. A result set that silently omits it looks identical to one where
+nothing matched, and *silence and success must never look alike*
+([ADR 0004](0004-a-node-and-how-it-joins.md)) — the rule this repository has now paid for twice.
+
+**A rule states how it is checked, and this one is checkable.** The check is the measurement that
+produced this record: search the mesh's memory for a phrase that appears only in a design document
+here, and require it back. That check fails today, deliberately, and passing it is what closes
+`04-ISSUES/006`.
+
+## References
+
+- [`04-ISSUES/006`](../04-ISSUES/006-hq-is-not-indexed-into-the-knowledge-base/00-report.md) —
+  the gap, the two measurements, and why closing it early was refused
+- [ADR 0019](0019-how-this-repository-works.md) — the promise this amends
+- [`README.md`](../README.md) — the objection, and the gap named where the claim used to sit

02-DECISIONS/0026-the-mesh-has-a-session-of-its-own.md

@@ -0,0 +1,130 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0004-a-node-and-how-it-joins.md
+---
+
+# 26. The mesh has a session of its own, and it is the node session's mechanism
+
+## Context
+
+[ADR 0004](0004-a-node-and-how-it-joins.md) gives every node a session: one per node, permanent,
+remembering across callers, its system prompt the node's engram, reachable over the broker like
+everything else. **Any node can message any node**, and that is called the one part of the system
+that is genuinely a mesh — symmetric, with no centre.
+
+**There is no way to address the mesh itself.** A question that spans machines — *what is running
+across all of this*, *which nodes are behind*, *why is it built this way* — has to be put to some
+node, which then asks the others. That works, and it makes a mesh-wide question **nobody's
+question**: every node answers it as a foreigner, from a position where the whole is not in view.
+
+**Three things independently arrived at the same missing piece.**
+
+[ADR 0025](0025-the-design-record-is-read-not-copied.md), taken hours before this one, commits to
+an agent that reads the design repository directly and answers into search. That agent has to
+exist, run somewhere, and be askable — and nothing in the record says what it is or where it
+lives.
+
+[`14-model-access.md`](../03-DESIGN/01-to-be/14-model-access.md) records, as a gap deliberately
+not half-built: *this worker uses that licence is a binding to an agent, not to a node* — and the
+provisions model has no consumer identity other than a node. A session that must be assigned a
+licence is exactly that consumer, and node sessions are already one.
+
+**And ADR 0004 never said how a session is set up.** It describes behaviour and stops: nothing
+states how a session starts, where its context lives, how the engram reaches it, or how a message
+off the broker becomes a prompt. There is no design document for it. That gap was invisible until
+something had to be built *like* a node session, because describing a second instance of a
+mechanism requires the mechanism to have been described once.
+
+## Considered Options
+
+1. **No mesh session; keep relaying through a node.** Costs nothing and works today. **Rejected.**
+   It leaves mesh-wide questions belonging to nobody, and it does not survive contact with
+   ADR 0025 — that agent still needs a home, so the thing gets built anyway, unnamed, as an
+   attachment to whichever node happened to host it.
+
+2. **A new kind of agent, built separately.** Purpose-built for the whole mesh. **Rejected.** It
+   would hold a session, a memory, a licence and broker plumbing — every one of which the node
+   session already has. Two implementations of one mechanism drift, and the vocabulary collision
+   that [ADR 0001](0001-mesh-brokers-nodes-host-agents-think.md) exists to undo began exactly this
+   way: two things that were nearly the same, built twice, until neither word meant one thing.
+
+3. **The same mechanism, started in a different context.** **Adopted.**
+
+## Decision
+
+**The mesh has one session, addressed as the mesh, and it is a node session in every respect but
+three.**
+
+| | |
+|---|---|
+| **the context it starts in** | the mesh's, not a machine's — this is the whole of what makes it different |
+| **its engram** | the mesh's system prompt, as a node's engram is that node's |
+| **its licence binding** | assigned in its own right, not inherited from the machine it runs on |
+
+Everything else is unchanged and deliberately so: it is permanent, it remembers, it is reachable
+over the broker, it holds its own tools, and switched off it still answers *I am switched off*
+rather than falling silent.
+
+**It runs on the node that holds the control plane** — not for convenience, but because that node
+is already the one place excepted from *compromise of a node is compromise of that node*
+(ADR 0004). An agent able to reach everything, placed anywhere else, creates a **second** such
+place. Putting it where the authority already sits concentrates nothing new.
+
+**It is an addition to per-node messaging and never a replacement.** Every node remains directly
+addressable. This is not a preference: ADR 0001 holds that losing the control plane costs *change,
+not operation*, and a mesh whose only conversational surface lives on that node would lose the
+ability to ask anything while every machine kept running perfectly. **The front door may not be
+the single point.**
+
+**It is not an employee** ([ADR 0003](0003-agents-are-persistent-employees.md)). Nobody hires it,
+it holds no task queue, it is never drained or reassigned. What it does with work that belongs
+somewhere else is **dispatch it** — to node sessions, or to workers — which is what a node session
+already does when asked something it does not have.
+
+**It is ADR 0025's reader.** The agent that reads the design repository and answers into search is
+this session, not a second one. One agent, one memory, one place to reach; two would both need
+that repository and would eventually disagree about what it says.
+
+**"One per node" is about address, not about process count.** ADR 0004's rule — *two and nothing
+decides which replies* — forbids ambiguity in who answers when a **node** is addressed. The mesh
+session answers when the **mesh** is addressed. The control-plane node therefore hosts two
+sessions and no ambiguity, and stating this here is what stops it reading as a contradiction
+later.
+
+## Consequences
+
+**The node session's setup must now be designed, and it never was.** This decision is expressed as
+*the same as a node session, elsewhere*, which is only meaningful once that mechanism is written
+down. The design document covering both is the immediate consequence of this record, not a
+follow-up to it.
+
+**A consumer that is not a machine stops being deferrable.** The licence binding above is the gap
+`14-model-access.md` names, and it now has two consumers rather than a hypothetical one. Until it
+exists, a session's model access can only be expressed as *this module on this machine*, which
+cannot say *this node's session uses the personal licence and the mesh's uses the company one* —
+the thing the binding is for.
+
+**Symmetry is preserved, and it is worth being precise about why.** ADR 0004's claim is about what
+a node can reach, and it is untouched: node-to-node messaging is unchanged, nothing is routed
+through the mesh session, and it is a participant rather than a hop. What arrives is a
+participant that happens to be the one a person usually addresses.
+
+**Availability degrades to inconvenience rather than to silence** — but only because of the
+addition rule above. If that rule is ever relaxed, this consequence inverts, and it inverts
+quietly: everything keeps working and nobody can ask about it.
+
+**The surface a person uses is not decided here.** That a board is a good place to talk to it is
+likely and is not this record's business; the session is reachable over the broker like everything
+else, and what puts a text box in front of it is a separate choice.
+
+## References
+
+- [ADR 0004](0004-a-node-and-how-it-joins.md) — the node session this extends
+- [ADR 0025](0025-the-design-record-is-read-not-copied.md) — the reader this session is
+- [ADR 0003](0003-agents-are-persistent-employees.md) — the vocabulary this is not
+- [`03-DESIGN/01-to-be/14-model-access.md`](../03-DESIGN/01-to-be/14-model-access.md) — *a
+  consumer that is not a machine*, the gap this makes concrete

02-DECISIONS/0027-a-provision-names-what-the-consumer-is-coupled-to.md

@@ -0,0 +1,109 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0009-modules-and-the-graph.md
+---
+
+# 27. A provision names what the consumer is coupled to, not the role it plays
+
+## Context
+
+Provisions are named after roles. The catalogue and every test fixture built so far say:
+
+```
+provides: database
+requires: database
+```
+
+**Nothing distinguishes one engine from another.** A module requiring `database` is satisfied by
+any module providing `database`, so a module written against PostgreSQL can be matched to a
+provider of Microsoft SQL Server, resolve as satisfied, deploy, and fail on its first query.
+
+**The mesh runs several engines** — PostgreSQL, Microsoft SQL Server, MariaDB, and others behind
+products that expose their own. This is not a hypothetical collision.
+
+**The failure is in the direction that hides.** Resolution *succeeds*. Nothing is refused, nothing
+is logged, and the breakage surfaces later as an error inside an application, on a machine, with
+nothing connecting it back to a match made elsewhere by something that thought it had done its
+job. **A wrong answer delivered confidently costs more than a refusal**, and the whole point of
+refusing on ambiguity ([ADR 0009](0009-modules-and-the-graph.md)) was to not do this.
+
+**How it got in:** every test written for the resolver had exactly one provider of each name, so no
+mismatch was expressible and none was caught. The fixtures agreed with the design. That is the same
+fault as [`04-ISSUES/005`](../04-ISSUES/005-pipeline-test-harness-unbuildable/00-report.md)'s
+imagined output and [`019`](../04-ISSUES/019-a-comment-asserting-a-fact-about-a-machine/00-report.md)'s
+unchecked comment, at the level of a name rather than a line.
+
+## Considered Options
+
+1. **Keep role names; let the operator assign correctly.** The mesh would refuse ambiguity when two
+   providers exist, so a person picks. **Rejected.** It makes correctness depend on somebody
+   knowing that the module they are assigning speaks a particular dialect — which is exactly the
+   knowledge the provisioning model exists to remove. And with one provider of each name, nothing
+   is ambiguous and nothing is asked.
+
+2. **A role name plus a `flavour:` or `engine:` qualifier**, matched as a second field.
+   **Rejected.** Two fields that must agree is a constraint the resolver has to enforce and a
+   manifest author has to remember, to express something one field already can. The name is the
+   contract; splitting it invites a requirement that names a role and forgets the qualifier, which
+   then matches everything again.
+
+3. **The name says what the consumer is coupled to.** **Adopted.**
+
+## Decision
+
+**A provision is named for the thing a consumer's code is written against.**
+
+```
+provides: postgres-database
+requires: postgres-database
+```
+
+**The test is whether the consumer can tell the difference.** If swapping the provider would break
+the consumer, the name must say which provider — because a match that breaks the consumer is not a
+match. If the consumer genuinely cannot tell, a role name is correct and better.
+
+| provision | | why |
+|---|---|---|
+| `postgres-database`, `mssql-database` | **specific** | applications are written against a dialect; a swap breaks them |
+| `route` | **role** | the consumer wants its name reachable and does not care what proxies it |
+| `resolver` | **role** | the consumer wants names to resolve |
+| `artifact-store` | **role** | the consumer fetches by digest over a protocol, and nothing else |
+
+**`database` is not a provision and may not be provided.** There is no context in which an
+application talks to a generic database: it talks to PostgreSQL or it talks to SQL Server. A name
+that cannot be true of any real consumer should not be expressible.
+
+**This is about coupling, not about products.** Two providers of `postgres-database` — a container
+on this node and a managed instance elsewhere — are interchangeable and *should* both match. What
+may not be interchangeable is what the consumer's queries are written in.
+
+## Consequences
+
+**Every manifest that names a database changes.** Doing this now costs a rename across a handful of
+examples. Doing it after modules are migrated costs it across all of them, plus every deployment
+that resolved against the old name.
+
+**Wrong requirements now fail loudly, and at the right moment.** A module requiring
+`postgres-database` where only `mssql-database` is provided is unsatisfiable, so it is **refused at
+resolution** with both names visible — rather than deployed and broken later. This is the property
+that was lost, restored.
+
+**Generic role names are still right, and the rule says when.** This does not push specificity
+everywhere; it puts it exactly where a consumer is coupled. Naming `route` after a particular proxy
+would be the same error in the other direction, and would prevent a swap that genuinely changes
+nothing.
+
+**It is checked, not merely stated** ([`00-META/how-we-build.md`](../00-META/how-we-build.md) §5).
+A manifest providing a name known to be engine-generic is refused, naming what to say instead.
+Without that, this record is a convention, and a convention is what the previous naming was.
+
+## References
+
+- [ADR 0009](0009-modules-and-the-graph.md) — provisions, and refusing on ambiguity
+- [`03-DESIGN/01-to-be/07-the-foundation.md`](../03-DESIGN/01-to-be/07-the-foundation.md) — *the
+  provisioning model uses databases, roles and schemas as PostgreSQL means them*, which is this
+  record's point made about the substrate before it was made about modules

02-DECISIONS/0028-the-substrate-supplies-the-control-plane-and-nothing-else.md

@@ -0,0 +1,118 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0006-the-substrate-and-the-control-plane.md
+---
+
+# 28. The substrate supplies the control plane and nothing else
+
+*Corrects one row of [ADR 0006](0006-the-substrate-and-the-control-plane.md) and makes explicit
+something it left unsaid. The rest of that record stands.*
+
+## Context
+
+ADR 0006 defines the substrate by a circularity: **what the control plane needs in order to run,
+and cannot ask itself for, because it is not running yet.** Two questions, and both must be
+answered *yes* for something to be substrate.
+
+Its membership table admits the object store on this line:
+
+| role | product | |
+|---|---|---|
+| object store | **MinIO** | it cannot grant itself a bucket |
+
+**That answers the second question and assumes the first.** It is true that a control plane cannot
+grant itself a bucket. Nothing establishes that it needs one.
+
+**It does not.** Verified 2026-08-31 against `mesh-control`: no S3 client, no bucket, no object
+storage of any kind outside comments. Artifacts reach nodes as content-addressed blobs in the OCI
+registry, and the code records the decision and its reasoning:
+
+> One store, and it is the registry the bootstrap already pulls from. An OCI registry is a
+> content-addressed blob store that happens to also understand images… The alternative considered
+> was a second store beside it — S3-shaped, buckets, signed URLs. It is the right answer for
+> objects that are *mutable*, or need per-reader access, or are not build output. None of that
+> describes a digest-pinned archive, and standing up a second service to hold one kind of
+> immutable blob means two things to run, two things to back up and two ways for an artifact to be
+> missing.
+
+**The row is inherited from the system being replaced**, where an object store distributed module
+tarballs. Here nothing does, and the row was never re-tested against the definition it sits under.
+
+**A second thing ADR 0006 never says:** whether a substrate service and a module of the same
+product are the same instance. It says the substrate is *not the control plane* and *not a place
+for logic*, and stops. The question is not idle — an application wanting a database, on a mesh
+whose substrate is already running PostgreSQL, has an obvious wrong answer available.
+
+## Considered Options
+
+1. **Leave the object store as substrate, unused.** Harmless-looking. **Rejected.** A membership
+   list that includes something nothing needs is a list that has stopped being derived from its
+   test, and the next member is admitted by precedent instead of argument. It also mandates that
+   every mesh run a service no mesh uses.
+
+2. **Applications share the substrate's instances.** One PostgreSQL, one of everything.
+   **Rejected**, below.
+
+3. **The substrate is exactly what the control plane consumes; everything else is a module.**
+   **Adopted.**
+
+## Decision
+
+**The object store is not substrate.** It fails the first half of the test: the control plane does
+not need one. An object store is an ordinary module, required through the module graph like
+anything else, and a module wanting one depends on a module providing one.
+
+**The substrate has four members, not five**: a relational store, a message bus, an image registry,
+and conditionally an identity provider. The registry stays — the control plane genuinely cannot
+deliver an artifact without somewhere to put it.
+
+**A substrate service and a module of the same product are different instances, and are not
+shared.** The mesh's own PostgreSQL and a PostgreSQL a workload was given are two servers, two
+containers, two lifecycles.
+
+Three reasons, and the first is the one that matters:
+
+**The substrate is not in the module graph.** It is raised from the pinned bundle the host carries,
+before any mesh exists to declare it. A workload depending on it would depend on something the
+graph cannot see, cannot rotate a credential for, and cannot move — which is every property the
+provisioning model exists to provide.
+
+**It would put workload data in the control plane's own store.** The mesh's contexts own their
+stores exclusively ([ADR 0008](0008-a-context-owns-its-store.md)). An application sharing that
+server can exhaust it, lock it, or fill its disk, and the failure is the control plane going down
+— which is the one failure that makes every other one harder to fix.
+
+**They are bounded differently.** The substrate is sized, backed up and upgraded as part of
+bootstrapping a mesh. A workload's database follows the workload — moved with it, destroyed with
+it, restored with it.
+
+## Consequences
+
+**Migrating an object store is ordinary module work**, not substrate work. It was previously going
+to be done as part of completing the substrate, which would have been the wrong shape and would
+have coupled every mesh to a service the mesh does not use.
+
+**A mesh with no workload needing one runs no object store at all.** That is the correct outcome
+and was not previously available.
+
+**Two PostgreSQL containers on a node that hosts both is expected**, not duplication to be
+optimised away. Anyone tidying them together should find this record first.
+
+**"Substrate by role and ordinary by delivery" loses one of its two members.** ADR 0006 uses that
+phrase of the object store and the registry — things that are substrate but provisioned once a
+control plane exists. It now describes the registry alone.
+
+**The definition is applied, not just stated.** Both halves of the circularity test are asked of
+each member, and *cannot grant itself one* is not sufficient on its own — it is true of almost any
+service, which is what made it possible to admit a member on that half alone.
+
+## References
+
+- [ADR 0006](0006-the-substrate-and-the-control-plane.md) — the definition, and the table this
+  corrects one row of
+- [ADR 0008](0008-a-context-owns-its-store.md) — a context owns its store exclusively
+- `mesh-control internal/builder/registry.go` — where artifacts go, and why not S3

02-DECISIONS/0029-a-network-is-a-shape-because-an-action-cannot-be-undone.md

@@ -0,0 +1,112 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0005-the-node-host.md
+---
+
+# 29. A network is a shape, because an action cannot be undone
+
+## Context
+
+**A module of several containers has no way to let them reach each other by name.** A container
+declaration carries a `network` field, and it only ever *joins* one that already exists — it was
+added so the control plane could reach the store and the broker on the machine it was raised on.
+Nothing in the vocabulary **creates** a network.
+
+Without one, containers on a machine share the runtime's default bridge, which gives addresses and
+no name resolution between them. So a module that is several containers can only be written by
+publishing ports onto the machine and pointing its own parts at the host — which puts a module's
+private wiring on the machine's own address space, where anything else on the machine can reach it
+and any other module can collide with it.
+
+**This is the gap a mail system meets and nothing else so far does**
+([`00-work-breakdown.md`](../03-DESIGN/01-to-be/00-work-breakdown.md) 3.3). It is being taken now
+rather than then, because 3.3 is the task most likely to send work back into the declaration
+language and the least useful place to discover it.
+
+**Adding a shape is not a small change, and the host says so** — the vocabulary is asserted
+against a stated number, with the reason written into the failure: *every addition widens what a
+compromised control plane can express, so a change here is a decision.* The host applies what it
+is told; the only bound on a hostile control plane is what the language can say
+([ADR 0004](0004-a-node-and-how-it-joins.md)).
+
+## Considered Options
+
+1. **An `action` that creates the network.** The vocabulary already has one, the bundle already
+   uses seven of them, and `docker network create` with a `verify` is exactly the shape an action
+   takes. Nothing would need adding. **Rejected**, on removal:
+
+   > An action has no footprint the host can undo — it ran, and whatever it did belongs to
+   > whatever it acted on.
+
+   A network made this way **leaks when the module is unassigned**, and the mesh cannot tell: the
+   record says an action ran, and there is nothing to reverse. Unassigning a module would leave a
+   network behind on every machine it was ever on, and the only way to find them would be to go
+   and look. *A resource the mesh can create and never clean up is one it should not create.*
+
+   There is a second reason, and it is the one that generalises: an action is opaque. **The mesh
+   cannot tell what an action did**, so a network created by one is not a thing the mesh knows
+   about — it cannot be reported, counted, or reasoned about, and a module could not require one.
+
+2. **Publish ports on the machine instead.** No new shape, and it works today. **Rejected.** It
+   makes a module's internal wiring part of the machine's address space: two modules that each
+   want a database on a fixed port collide, and anything else on the machine can reach what was
+   meant to be private. It also makes the module's manifest depend on what else is installed,
+   which is the thing provisioning exists to remove.
+
+3. **`network` as a ninth shape.** **Adopted.**
+
+## Decision
+
+**`network` joins the vocabulary, and the vocabulary is nine shapes.**
+
+```
+{"id": "internal", "type": "network", "name": "mail"}
+```
+
+**A name and nothing else.** Not a driver, a subnet, an address range or a gateway: every one of
+those is a thing a module would have to know about the machine it lands on, and a module that
+names a subnet is a module that collides with whatever else chose the same one. The runtime picks;
+the mesh names.
+
+**It is created if absent and removed when no longer declared** — an ordinary shape, with the same
+lifecycle as a directory. That is the whole reason it is a shape.
+
+**Declared before the containers that join it.** Resources are applied in the order the module
+wrote them, and orphans are removed in **reverse** — so a network written first is created first
+and removed last, after the containers attached to it are gone. This is not a new rule; it is the
+existing one, and it happens to be exactly right here. A network written *after* its containers
+would fail to remove while they still hold it, and that failure is reported rather than silent.
+
+**What it does not do:** it does not reach across machines. A network is one machine's, like
+everything else the host applies. Modules on different machines reach each other over the private
+network the mesh already provides ([ADR 0007](0007-connectivity.md)), and a shape that tried to
+span machines would be a second overlay with a worse contract.
+
+## Consequences
+
+**The vocabulary is nine, and the count moves with a record.** The test that asserts it names this
+one, so the next person to change it finds the argument rather than a number to edit.
+
+**A compromised control plane can now create and destroy networks on a machine.** Stated plainly
+because that is the cost, and the bound is the point: it can create a named network and remove
+one, and it can do neither to anything it did not declare. It cannot inspect, attach to, or
+reroute what is already there — those would be different shapes, and are not being added.
+
+**A multi-container module becomes expressible**, which unblocks 3.3 and, less obviously, makes
+several smaller modules simpler: anything that is a service plus a sidecar currently has to
+publish a port to talk to itself.
+
+**Nothing is required to use it.** A module of one container declares no network and joins none,
+exactly as now. The substrate keeps using `host`, which is a runtime-provided network and not one
+the mesh creates.
+
+## References
+
+- [ADR 0005](0005-the-node-host.md) — the host's vocabulary, and why each shape is a decision
+- [ADR 0004](0004-a-node-and-how-it-joins.md) — what may be pushed is bounded by form, not by trust
+- [`03-DESIGN/01-to-be/00-work-breakdown.md`](../03-DESIGN/01-to-be/00-work-breakdown.md) — 1.3,
+  and the mail system at 3.3 that this is for

02-DECISIONS/0030-data-outlives-the-mesh-that-declared-it.md

@@ -0,0 +1,96 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0005-the-node-host.md
+---
+
+# 30. Data outlives the mesh that declared it
+
+## Context
+
+**The conversion runs on live services holding real data**, and starts on the node that holds all
+of it. Identity, mail, everything. The requirement stated plainly: a data directory may be
+*moved*, and may never be *lost*.
+
+**The host deleted them.** A directory that stopped being declared was an orphan, and an orphan
+directory was removed with `os.RemoveAll` — everything under it — while the report said
+`removed`. A module unassigned took its database's files with it, and nothing anywhere said what
+had been in there.
+
+Reproduced before it was fixed: assign a module, let a service write into its directory, unassign
+the module, and the file is gone.
+
+**A directory stops being declared for ordinary reasons**, which is what makes this sharp rather
+than theoretical. A module unassigned from a node. A manifest edited to move a data folder — the
+exact operation the conversion needs. A resource renamed. A typo. **Every one of those is a normal
+day's work, and every one of them was destructive.**
+
+**The removal order was already right, and that is what makes a fix possible.** Everything the
+mesh puts inside a directory is itself a declared resource, and orphans are removed in reverse
+declaration order — so by the time a directory is reached, what the mesh wrote there is already
+gone. Anything still present was put there by something else.
+
+## Considered Options
+
+1. **A `keep` flag on the directory.** A module declares which of its directories hold data, and
+   the host leaves those. **Rejected.** It is safe only when somebody remembered, and the failure
+   of forgetting is total and silent. A rule that protects data only when it was asked to is not
+   a rule about data, it is a rule about attentiveness — and this is the one place in the system
+   where being wrong does not recover.
+
+2. **Never remove a directory.** Simple and unarguably safe. **Rejected**, narrowly: every module
+   ever assigned would leave its directories behind for ever, and a machine that accumulates
+   things nobody can account for is one where nobody can tell what is still in use. The clean-up
+   that is genuinely the mesh's is worth keeping.
+
+3. **Remove a directory only when it is empty.** **Adopted.**
+
+## Decision
+
+**A directory that still holds anything is kept, and the mesh says so.** An empty one is removed.
+
+**This is the host's existing line applied to the one shape where getting it wrong is
+unrecoverable** — *it removes what it made and leaves what it merely configured*
+([ADR 0005](0005-the-node-host.md)). An empty directory is what the host made. A full one is not.
+
+**No flag, no declaration, nothing to remember.** Emptiness is the test, and it is derived from
+the removal order rather than asserted: the mesh's own contents are gone by then, so what remains
+is by definition something nobody declared.
+
+**It is reported, not silent.** The outcome is `kept`, naming how many items are inside and saying
+they are for a person to deal with. A directory quietly left behind is how a machine accumulates
+things nobody can account for — which is the objection to option 2, and it is answered by saying
+so rather than by deleting.
+
+**Files are unchanged.** A declared file is the mesh's own — it wrote it, it owns it, and losing a
+configuration file is not the failure this is about. The distinction is deliberate: **directories
+hold what other things produced; files are what the mesh itself put there.**
+
+## Consequences
+
+**Moving a data directory is now safe by default.** The manifest changes, the old path stops being
+declared, and the data stays where it is until somebody has looked at it. That was the operation
+most likely to destroy something during the conversion, and it is now the operation that does the
+least.
+
+**Unassigning a module leaves its data.** Correct, and it means unassignment is no longer a way to
+clean up — removing data is a person's act, done knowingly. Given what unassignment did before,
+that is the trade being made and it is the right way round.
+
+**A machine can accumulate directories nobody removed.** Accepted, and mitigated by saying so
+every time rather than by a periodic sweep. A sweep would be the deletion this record exists to
+prevent, on a timer, with nobody watching.
+
+**It is not a backup, and must not be mistaken for one.** This stops the mesh destroying data. It
+does nothing about a disk, a mistaken `rm`, or a service corrupting its own store. The conversion
+still needs backups taken and **restored** before anything is moved — a backup nobody has restored
+is a belief, not a copy.
+
+## References
+
+- [ADR 0005](0005-the-node-host.md) — the host removes what it made
+- [`03-DESIGN/01-to-be/00-work-breakdown.md`](../03-DESIGN/01-to-be/00-work-breakdown.md) — the
+  conversion this was found by planning

02-DECISIONS/0031-the-control-plane-authenticates-nobody.md

@@ -0,0 +1,72 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0006-the-substrate-and-the-control-plane.md
+---
+
+# 31. The control plane authenticates nobody, so identity is a module
+
+## Context
+
+[ADR 0006](0006-the-substrate-and-the-control-plane.md) left one member of the substrate
+conditional, and said exactly why:
+
+| role | product | |
+|---|---|---|
+| identity provider | — | **conditional**: substrate only if the control plane delegates authentication, which is undecided |
+
+[`07-the-foundation.md`](../03-DESIGN/01-to-be/07-the-foundation.md) carried it as an open question —
+*whether identity is the fifth* — noting it followed from a decision nobody had taken.
+
+**The decision is taken: the control plane does not delegate authentication.** There is no mesh
+identity provider.
+
+**Nothing in the mesh's own machinery ever needed one.** A node proves itself with a keypair it
+generated, over a broker account issued at enrolment
+([ADR 0004](0004-a-node-and-how-it-joins.md)). Declarations are verified by signature. None of
+that touches an identity provider, and the conditional was never about machines — it was only ever
+about whether a *person* signing in to a mesh surface would be authenticated by something else.
+
+## Decision
+
+**Identity is a module**, like the mail system and the forge. It runs *on* the mesh, not *of* it
+([ADR 0001](0001-mesh-brokers-nodes-host-agents-think.md)) — a provider other modules require,
+which is the ordinary shape and needs nothing new to express.
+
+**So the substrate is three, and no longer conditional**: a relational store, a message bus, and
+an image registry. Together with
+[ADR 0028](0028-the-substrate-supplies-the-control-plane-and-nothing-else.md), which removed the
+object store, the list is settled and every member is there for the same reason — the control
+plane needs it and cannot ask itself for it.
+
+**A mesh that wants no identity provider runs none.** That is now expressible, and was not while
+it sat in the substrate as a maybe.
+
+## Consequences
+
+**The last open question about substrate membership is closed.** Both halves of ADR 0006's test
+now have an answer for every candidate, and the answer for identity is *the control plane does not
+need it*.
+
+**It does not settle how a person signs in to a mesh surface**, and that is deliberately left
+open. What is settled is that whatever answers it is not part of what must exist before the mesh
+does — so it can be decided late, changed, or replaced, which is precisely what being substrate
+would have prevented.
+
+**It becomes a real test of the module graph.** An identity provider is a module that *other
+modules require* — the object store already consumes it — so it exercises the provider chain more
+seriously than anything ported so far, where the provider was written alongside its consumer.
+
+**Ordering follows from it rather than from preference.** Anything requiring identity has to move
+after it, which is a dependency the graph can state rather than something a person has to
+remember.
+
+## References
+
+- [ADR 0006](0006-the-substrate-and-the-control-plane.md) — the conditional this closes
+- [ADR 0028](0028-the-substrate-supplies-the-control-plane-and-nothing-else.md) — the other member
+  removed, and the test applied properly
+- [ADR 0004](0004-a-node-and-how-it-joins.md) — how a node proves itself, which needs none of this

02-DECISIONS/0032-the-local-account-owns-the-mesh.md

@@ -0,0 +1,79 @@
+---
+topic: how we work
+status: superseded
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0031-the-control-plane-authenticates-nobody.md
+superseded-by: 02-DECISIONS/0034-the-local-account-owns-the-mesh.md
+---
+
+# 32. The local account owns the mesh; a surface delegates to a module
+
+## Context
+
+[ADR 0031](0031-the-control-plane-authenticates-nobody.md) settled that the control plane
+authenticates nobody, and deliberately left one thing open: **how a person signing in to a mesh
+surface is authenticated.** This answers it, and answers a question 0031 did not ask — *who owns
+the mesh at all.*
+
+**There was no answer, and the absence was invisible** because every operation so far has been run
+by the person sitting at the machine. Nothing had to say whether that was the design or the
+circumstance.
+
+## Decision
+
+**The account that installed the host owns the mesh on that node.** Authority is a local login,
+and there is nothing else to hold.
+
+**No mesh user model.** No accounts, no roles, no grants, nothing to administer. A person with a
+shell on a node can do anything the mesh can do there, because that is already true and pretending
+otherwise would be a boundary that does not exist.
+
+**This follows from what was already decided rather than adding to it.**
+[ADR 0004](0004-a-node-and-how-it-joins.md) says there is no authorisation between nodes — every
+node is the operator's own, so a message from one is a message from them, and *the mesh boundary
+is therefore the security boundary*. A user model inside that boundary would guard nothing: anyone
+who could be stopped by it could equally read the node's key off the disk.
+
+**The board is different, and the difference is the network.** A surface reachable by a browser
+has to know who is asking, because the people reaching it are not, by construction, people with a
+shell on the machine. **So the board delegates to an OAuth provider** — which is a module.
+
+## What this does not change
+
+**The identity provider is still not substrate** (ADR 0031). A *surface* delegating
+authentication is not *the control plane* delegating it. The control plane runs, applies
+declarations and reaches nodes with no identity provider in existence; only the board needs one,
+and only to decide whose browser it is talking to.
+
+The test is unchanged and still answers no: *does the control plane need it in order to run?*
+
+## Consequences
+
+**The board depends on a module, and says so.** An ordinary edge in the graph, which means the
+board cannot come up before the provider it authenticates against — stated as a dependency rather
+than discovered as an outage.
+
+**Moving the identity provider takes the board with it.** During that module's own conversion the
+board is unavailable, and that is acceptable: it is a surface, nothing depends on it, and a brief
+interruption is the trade already accepted everywhere else. Nothing that keeps a service serving
+goes through it.
+
+**Anyone with a shell on a node has full authority there.** Written down rather than left implied,
+because it is the sentence that decides who gets an account on a machine. The protection is the
+machine's own login, and the overlay that keeps the machine unreachable from outside
+([ADR 0007](0007-connectivity.md)).
+
+**A node cannot be operated by somebody without a login on it.** Deliberate, and the cost of
+having no user model: there is no way to give a person authority over one node without giving them
+a shell there. If that is ever wanted, it is a new decision and not a gap in this one.
+
+## References
+
+- [ADR 0031](0031-the-control-plane-authenticates-nobody.md) — the control plane authenticates
+  nobody; this answers what it left open
+- [ADR 0004](0004-a-node-and-how-it-joins.md) — no authorisation between nodes, and why the mesh
+  boundary is the security boundary
+- [`03-DESIGN/01-to-be/11-a-board.md`](../03-DESIGN/01-to-be/11-a-board.md) — the surface this is
+  about

02-DECISIONS/0033-the-substrate-is-a-store-and-a-broker.md

@@ -0,0 +1,91 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0028-the-substrate-supplies-the-control-plane-and-nothing-else.md
+---
+
+# 33. The substrate is a store and a broker
+
+## Context
+
+Third correction to one table in one day, all found the same way: by asking whether **both** halves
+of the substrate test were actually answered for a given member, or only the second.
+
+The test ([ADR 0006](0006-the-substrate-and-the-control-plane.md)) is *what the control plane needs
+in order to run, and cannot ask itself for, because it is not running yet.* ADR 0006 admits the
+image registry on this line:
+
+| role | product | |
+|---|---|---|
+| image registry | **an OCI registry** | it cannot grant itself a repository |
+
+**That is the second half again.** It is true that a control plane cannot grant itself a
+repository. Nothing establishes that it needs one *in order to run*.
+
+**Counted rather than argued.** `substrate-first-node.lock` — the only bundle there is, and what a
+first node actually becomes — raises twelve resources, and no registry is among them:
+
+```
+container runtime · the store · one database per context · the schemas
+· the broker's certificate · the broker · the control plane
+```
+
+The registry arrives afterwards, as an ordinary module the mesh assigns. That is what the lab
+asserts, in those words: *the mesh runs its own artifact store.*
+
+**ADR 0006 half-said this already**, calling the registry *substrate by role and ordinary by
+delivery, provisioned once there is a control plane to do it.* A member that is provisioned by the
+thing it supposedly precedes is not a member; the phrase was carrying a contradiction rather than
+resolving one.
+
+**The registry is a closer call than the object store, and the difference is worth keeping.** The
+control plane never touches an object store at all — no client, no bucket, ever
+([ADR 0028](0028-the-substrate-supplies-the-control-plane-and-nothing-else.md)). It genuinely
+*uses* the registry: the builder pushes to it, hosts pull from it, and nothing reaches a machine
+without it. **So the registry is a real dependency of the mesh operating, and not of the control
+plane starting** — and it is the second that the word substrate means.
+
+## Decision
+
+**The substrate is two things: a relational store and a message bus.** Both are in the bundle,
+both must exist before the control plane's first instruction, and neither can be asked for.
+
+**The registry is an ordinary module.** The mesh cannot deliver anything without one, and it
+installs one the way it installs everything else. The first node's chicken-and-egg is already
+solved and needs nothing from this list: it fetches upstream images directly, then runs a registry
+of the mesh's own.
+
+**The test is applied to both columns, every time.** *Cannot grant itself one* is true of almost
+any service and settles nothing on its own. It is what admitted the object store, and then the
+registry, and both were removed by asking the other question.
+
+## Consequences
+
+**The substrate is now exactly what the bundle raises**, which is the strongest form this list can
+take: it can be checked by counting rather than by reading an argument. A member that is not in
+the bundle is not substrate, and the two statements cannot drift apart.
+
+**A mesh that builds nothing still needs a registry** — to receive anything at all — but it needs
+it as a module, on its own schedule, replaceable. That was already true and was obscured by the
+list.
+
+**The word may now be doing too little work.** "Substrate" for *a database and a broker* is a term
+of art for two things everybody can name. Renaming is not taken here and is worth considering
+separately; what this record fixes is the membership, not the vocabulary.
+
+**Three removals from one table in one day is itself the finding.** Each member was admitted on the
+half of the test that is easy to answer, and the design read plausibly throughout. The rule that
+comes out of it is not about substrates: **a test with two conditions is a test only when both are
+asked.**
+
+## References
+
+- [ADR 0006](0006-the-substrate-and-the-control-plane.md) — the definition, and the table this
+  corrects a second row of
+- [ADR 0028](0028-the-substrate-supplies-the-control-plane-and-nothing-else.md) — the object
+  store, removed for the same reason
+- [ADR 0031](0031-the-control-plane-authenticates-nobody.md) — identity, which was conditional and
+  is now a module

02-DECISIONS/0034-the-local-account-owns-the-mesh.md

@@ -0,0 +1,81 @@
+---
+topic: how we work
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+supersedes: 02-DECISIONS/0032-the-local-account-owns-the-mesh.md
+---
+
+# 34. The local account owns the mesh, and a web application's login is not that
+
+*Supersedes [ADR 0032](0032-the-local-account-owns-the-mesh.md), which decided the right thing and
+described it wrongly. The decision below is unchanged; what it said about the board was an
+invention.*
+
+## Context
+
+ADR 0032 answered *who owns the mesh* — the account that installed the host — and then framed the
+board as **a surface that delegates authentication**, a category it made up for the occasion. It
+does not need one.
+
+**The board is a web application.** It has a login, provided by the identity module, in the way
+every web application has a login. That is a fact about an application, not a property of the
+mesh, and giving it a name in the mesh's vocabulary implied a relationship that is not there.
+
+The cost of the invented category was not cosmetic. It made the identity module look like part of
+the mesh's own authority — something the mesh *depends on* to know who anybody is — when the truth
+is that the mesh knows nothing about people at all, and one of the applications running on it has
+a login.
+
+## Decision
+
+**The account that installed the host owns the mesh on that node.** Authority is a local login.
+There is nothing else to hold, no user model, no roles, and nothing to administer.
+
+**This follows from what was already decided.**
+[ADR 0004](0004-a-node-and-how-it-joins.md) says there is no authorisation between nodes — every
+node is the operator's own, so a message from one is a message from them, and *the mesh boundary
+is the security boundary.* A user model inside that boundary would guard nothing: anyone it could
+stop could read the node's key off the disk.
+
+**A web application's login is its own business.** The board authenticates its users through the
+identity module. So might anything else the mesh runs. **None of that is mesh authority**, and the
+mesh does not learn who anybody is from it.
+
+## The line this draws, which is the reason to write it down
+
+**Signing in to an application must not, on its own, become authority over the mesh.**
+
+Today it cannot: the board reads and does not act
+([`11-a-board.md`](../03-DESIGN/01-to-be/11-a-board.md) — *not the way to change things*). Looking
+at a page tells you what is true and changes nothing.
+
+**The moment the board can assign a module, whoever it lets in has mesh authority** — and it would
+arrive as a feature rather than as a decision. That is the failure this record exists to make
+visible, because it is the kind that is only obvious afterwards.
+
+So: **a surface that can change the mesh is a change to who owns the mesh**, and is taken as one.
+Not forbidden — wanting to manage nodes from a browser is reasonable — but not something that
+turns up in a pull request titled *add assign button*.
+
+## Consequences
+
+**The identity module is not special.** Not substrate ([ADR 0031](0031-the-control-plane-authenticates-nobody.md)),
+not part of the mesh's authority, and nothing about the mesh stops working when it is down. Some
+applications cannot be logged into, which is what it means for an application's login provider to
+be unavailable.
+
+**Anyone with a shell on a node has full authority there.** Unchanged from ADR 0032, and still the
+sentence that decides who gets an account on a machine. The protection is the machine's own login
+and the overlay that keeps it unreachable from outside ([ADR 0007](0007-connectivity.md)).
+
+**A node cannot be operated by somebody without a login on it.** The cost of having no user model.
+If that is ever wanted, the paragraph above says what it costs.
+
+## References
+
+- [ADR 0032](0032-the-local-account-owns-the-mesh.md) — superseded; same decision, invented category
+- [ADR 0004](0004-a-node-and-how-it-joins.md) — the mesh boundary is the security boundary
+- [ADR 0031](0031-the-control-plane-authenticates-nobody.md) — the control plane authenticates
+  nobody

02-DECISIONS/0035-one-implementation-several-surfaces.md

@@ -0,0 +1,115 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0034-the-local-account-owns-the-mesh.md
+---
+
+# 35. One implementation, several surfaces, and what that costs
+
+## Context
+
+The mesh is operated from a command line today. It needs to be operable from a browser and from a
+model's tools as well, and the three must not be three different systems.
+
+**The pattern is already in the code and unnamed.** `board` serves HTTP by calling the same
+functions the CLI calls; it holds nothing and decides nothing. What follows makes that the rule
+rather than a property of one command.
+
+**The board is a presentation layer over the control plane.** Not an application beside it holding
+a database credential — the thing that shows what the control plane knows, and asks it to do what
+a person asked for.
+
+## Decision
+
+**The logic lives once, in the context that owns it. A surface is an adapter with no decisions in
+it.**
+
+| surface | for |
+|---|---|
+| **command line** | a person on a machine, and the recovery path below |
+| **HTTP** | the board, and anything else that speaks to the mesh over a network |
+| **model tools** | an agent asking the mesh to do something |
+
+**Every surface refuses identically, because the refusal is not in the surface.** An assignment
+that cannot be satisfied is refused by the same resolution whichever way it arrived. The moment a
+surface can accept something another would reject, the mesh has two answers to one question and
+people learn which to trust.
+
+**Reading and doing are both exposed.** The HTTP surface is not read-only: managing the mesh from
+a browser is the point. This takes the decision
+[ADR 0034](0034-the-local-account-owns-the-mesh.md) said had to be taken deliberately —
+**a browser login now carries authority over the mesh** — and takes it knowingly rather than
+letting it arrive with a feature.
+
+**The networked surfaces authenticate through an OAuth2 identity provider.** Named by protocol
+rather than by product, like every other dependency the mesh takes — AMQP for the bus, S3 for an
+object store, OCI for the registry
+([ADR 0006](0006-the-substrate-and-the-control-plane.md)). What fills the role today is a module
+running Keycloak; what the control plane knows is that it validates a token against a provider
+speaking OAuth2, and replacing that provider is a migration rather than a redesign.
+
+**The command line does not authenticate at all**: it is already behind the machine's own login,
+which is what owns the mesh (ADR 0034).
+
+## What this is not: a kernel every module imports
+
+**The shared library is the failure this project was started over**, and the difference has to be
+stated or it will be rebuilt. The old one is 155 files and 34,636 lines *containing code from
+every context* — work-domain logic sitting in the kernel every module imports, each piece landing
+there to avoid a cycle between two modules that both needed it.
+
+**Shared surfaces are not a shared library.** What is shared here is that three adapters call the
+same functions. Those functions stay in the context that owns them — provisioning's logic in
+provisioning, identity's in identity — and no module imports another's. A surface may call many
+contexts; a context still may not reach into another's store
+([ADR 0008](0008-a-context-owns-its-store.md)).
+
+The test, when something is about to be put "somewhere shared": *does this belong to a context, or
+does it only belong to the surface?* If it belongs to a context it goes there, even if two
+surfaces want it.
+
+## The loop this creates, and the way out
+
+**The control plane's networked surfaces will depend on a module the control plane assigns.**
+An identity provider is an ordinary module ([ADR 0031](0031-the-control-plane-authenticates-nobody.md)). When
+it is down, or being migrated, or misconfigured, the HTTP and tool surfaces cannot authenticate
+anybody — including the person trying to fix it.
+
+**The command line is the way out, and it is why local ownership matters more rather than less.**
+It authenticates through nothing, needs no network, and is available on the machine to the account
+that owns the mesh. **A mesh must always be operable by somebody standing at it.**
+
+So the rule: **no capability exists only behind an authenticated surface.** Anything the board can
+do, the command line can do. That is not a courtesy to CLI users; it is the recovery path, and a
+capability that exists only over HTTP is one that disappears exactly when identity does.
+
+## Consequences
+
+**Identity is still not substrate**, and the test still answers no: the control plane runs, applies
+declarations and reaches nodes with no identity provider in existence
+([ADR 0033](0033-the-substrate-is-a-store-and-a-broker.md)). What is unavailable without it is two
+surfaces, not the mesh.
+
+**Whoever the identity provider admits has authority over the mesh.** That is now a real perimeter
+with real consequences, where before it guarded a page that only read. Who may log in, and to
+which realm, becomes a decision about the mesh rather than about an application.
+
+**A surface must not grow an opinion.** The likely erosion is a validation added to the board
+because it was quicker there — and then the CLI accepts something the board rejects, or worse the
+reverse. Adapters hold no decisions.
+
+**Three surfaces over one implementation is a cost paid three times if it is not one
+implementation.** The reason to write this down now is that the second surface is the cheapest
+moment to get it right, and the third is where the drift usually starts.
+
+## References
+
+- [ADR 0034](0034-the-local-account-owns-the-mesh.md) — the local account owns the mesh, and the
+  line this record deliberately crosses
+- [ADR 0001](0001-mesh-brokers-nodes-host-agents-think.md) — the shared library this must not
+  become
+- [ADR 0008](0008-a-context-owns-its-store.md) — a context owns its store, which a surface does
+  not change

02-DECISIONS/0036-bootstrap-ends-at-a-usable-mesh.md

@@ -0,0 +1,99 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-08-31
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0035-one-implementation-several-surfaces.md
+---
+
+# 36. Bootstrap ends at a usable mesh, and the first credential comes from a person
+
+## Context
+
+Bootstrap currently ends when the control plane starts
+([`07-the-foundation.md`](../03-DESIGN/01-to-be/07-the-foundation.md)). That is a mesh that runs and
+cannot yet be used by anybody who is not standing at the machine: the networked surfaces need an
+OAuth2 identity provider ([ADR 0035](0035-one-implementation-several-surfaces.md)), the provider is
+a module, and no module has been assigned.
+
+**So bootstrap should go further** — through the identity provider and the first login — and stop
+at a mesh somebody can actually use.
+
+**One thing in the way, and it is not incidental.** The mesh has never held a readable secret. The
+sealing code says what it does and why:
+
+> Make generates a secret and seals it to both ends, **keeping no readable copy.**
+
+An initial administrator's credential is the first value a **person must read**. Everything else
+the mesh generates is something no human ever sees, and everything a human provides is something
+the mesh immediately stops being able to read.
+
+## Considered Options
+
+1. **The mesh generates it and prints it once**, to the terminal of whoever ran the bootstrap.
+   Convenient, and needs no prompt. **Rejected.** It would give the control plane a plaintext
+   secret for the first time — briefly, and only to one terminal, but the capability would then
+   exist. *An exception made for one case does not stay one*: the next credential that is awkward
+   to supply gets printed too, and the property that a copy of the mesh's database is a copy of
+   nothing stops being checkable by reading the code.
+
+2. **No password: a one-time link that lets the operator set their own.** The nicest to use.
+   **Rejected for now** — it needs a mechanism that does not exist, and the thing it improves is
+   one prompt, once, on a new mesh.
+
+3. **The operator supplies it.** **Adopted.**
+
+## Decision
+
+**Bootstrap runs to a usable mesh**: the substrate, the control plane, the identity provider as an
+ordinary module, its realm and client provisioned, an administrator able to log in, and the
+networked surfaces available.
+
+**The administrator's credential is supplied by the person doing the bootstrap**, on standard
+input and not echoed — the path that already exists for a model-access key. The mesh seals it and
+cannot read it afterwards.
+
+**What is created is an account in the identity provider, not a user of the mesh.** The mesh still
+has no user model and gains none here ([ADR 0034](0034-the-local-account-owns-the-mesh.md)). What
+this produces is the first login for the applications that have one.
+
+**The provisioning is ordinary.** A realm, a client and a first account are what an identity
+module's provisioner makes from what the mesh granted it — the same shape as a database and a
+bucket, which are built and proven.
+
+**The surfaces arrive when their dependency does.** The command API is not started with the
+control plane and then broken until identity exists; it becomes available once it can authenticate,
+the way anything else waits for a provider.
+
+## Consequences
+
+**The mesh still never holds a readable secret**, and that sentence needs no exception clause.
+That is the whole reason for the prompt.
+
+**An unattended bootstrap is still possible, and the value still comes from outside.** Automation
+supplying the credential is the operator supplying it. What is refused is the *mesh inventing*
+one — so an unattended bootstrap with no credential provided produces a mesh with no
+administrator, which is correct rather than broken.
+
+**Bootstrap gains an interactive step**, and it is the only one. Worth stating because a bootstrap
+that cannot run without a person is a real constraint on how a node is stood up, and this is
+deliberate rather than an oversight.
+
+**The identity provider is still not substrate.** It is assigned by the control plane, so it comes
+after it, and a thing that comes after cannot be a thing that must exist before
+([ADR 0033](0033-the-substrate-is-a-store-and-a-broker.md)). Bootstrap running through it does not
+move it: bootstrap is a sequence, the substrate is a dependency.
+
+**And the recovery path is unchanged.** When the identity provider is broken later — which is the
+failure that matters, not the one at first start — the command line still works, because it
+authenticates through nothing (ADR 0035).
+
+## References
+
+- [ADR 0035](0035-one-implementation-several-surfaces.md) — the surfaces, and why the command line
+  must keep working
+- [ADR 0034](0034-the-local-account-owns-the-mesh.md) — the local account owns the mesh; this adds
+  no user model
+- [ADR 0033](0033-the-substrate-is-a-store-and-a-broker.md) — what must exist before the control
+  plane, which this does not change

02-DECISIONS/0037-where-a-module-lives.md

@@ -0,0 +1,103 @@
+---
+topic: building it
+status: proposed
+date: 2026-09-01
+deciders: jochen
+reconstructed: false
+rests-on: 02-DECISIONS/0009-modules-and-the-graph.md
+---
+
+# 37. Where a module lives
+
+## The question
+
+The mesh's own module descriptions currently sit in `examples/` inside the control plane, beside
+the small programs that hand out logins. That was fine while there were three of them. It is
+wrong now, and the name is doing active harm: everything in `examples/` reads as a sketch, and one
+of them shipped naming a container image that nothing in the repository builds. A directory called
+*the catalogue* would have made *does this actually work* the obvious question to ask of it.
+
+So: **one repository holding the modules we ship?** And if so, where does everything that is not
+ours go?
+
+## What a module actually is, counted
+
+The system being replaced has **126 modules** on its main branch. The shape of them is the whole
+argument, so it is measured rather than assumed:
+
+| | count | what it is |
+|---|---|---|
+| **the module is software** | 47 | its own source tree lives inside the module — a daemon, a service, a library |
+| **helper scripts only** | 44 | no application of its own; scripts it runs at install time or offers to an agent |
+| **a description and nothing else** | 35 | a package to install and some files to write |
+
+**Two thirds of modules contain code.** The largest is a shared library of 182 source files. A
+speech-capture module carries a complete daemon — audio capture, mixing, transcription, a model
+runner. Treating a module as *a description of something else* is true of barely a quarter of them.
+
+That kills the simplest answer. A catalogue cannot be "a folder of manifests" when most modules
+are programs.
+
+## The four kinds, which want different homes
+
+**1. What the mesh is made of.** The control plane, the host, the shared library, the board.
+These are not modules that happen to be ours; they are the mesh, expressed as modules so it can
+install itself. They belong in the repositories that build them, which already exist.
+
+**2. Something the world made, that we describe.** A forge, a mail system, an identity provider,
+a media server. Nobody upstream ships a description; somebody has to write one, and it is the same
+description for everybody who runs it. **This is what a catalogue is for.** It is also where the
+small programs that create accounts belong, because such a program is part of describing that
+service, not part of the mesh.
+
+**3. Something we wrote, that runs somewhere.** An application, a site, a side project. The
+description belongs **with the code, at the root of its own repository**, because the two change in
+the same commit. A repository that gains an environment variable and a description that gains it
+elsewhere will drift, and there is no mechanism that could stop it. This is already how it works
+and it should stay that way.
+
+**4. A package and some files.** A tool, a font, a shell. Thirty-five of these, and each is a few
+lines. The catalogue.
+
+## The proposal
+
+**A `mesh-catalog` repository** holding kinds 2 and 4: descriptions of software we did not write,
+and the programs that provision it. Not kind 1, which is the mesh itself. Not kind 3, which lives
+with its own code.
+
+**The mesh's list of modules is not this repository.** It is a table in the control plane, filled
+by adding a description to a running mesh. The catalogue is a *source* to add from — one of
+several, and the mesh already records which: every module carries where it came from, the branch
+followed there, and the commit its description was read at. **Nothing needs inventing to support
+modules from anywhere**; a repository of our own is simply the source we curate.
+
+**A description is checked by the tool, not by a test that imports the tool.** Today a test in the
+control plane parses the example manifests by reaching into the control plane's internals, and
+another reads the control plane's own build file to check every image a module names can be built.
+Two jobs tangled. A `module check` command on the control plane's binary would let the catalogue
+hold data validated from outside, and would give the same check to somebody describing their own
+application in their own repository — which is the case that matters most and currently has no
+check at all.
+
+## What this costs, and the argument against
+
+**It is early.** Ten modules exist, four of them ours. Moving ten files is a morning; moving a
+hundred is a week — but the hundred is not here yet, and splitting now adds a second repository to
+release across before there is anything to release.
+
+The counter is that the tangle is already producing faults rather than merely threatening to. A
+manifest naming an unbuildable image, and a test reading a build file two directories up, are both
+symptoms of one repository doing two jobs. And the moment the first module is adopted on a real
+machine, the descriptions stop being examples and become the thing deployments come from. **That
+is the moment this becomes urgent, and it is close.**
+
+## What it does not settle
+
+**Where a provisioning program's image is published**, and how a description pins it. A description
+names an image by digest; the image is built from the catalogue; the catalogue must therefore both
+produce an image and refer to it, which is the same knot the bootstrap has and solves by writing
+the digest down after building.
+
+**Whether kind 4 deserves a module at all.** Thirty-five descriptions that say *install this and
+write these files* may be better as one module with settings than as thirty-five modules. Left
+open deliberately; it is a question about the shape of the catalogue, not about whether to have one.

02-DECISIONS/0038-the-mesh-assigns-the-port.md

@@ -0,0 +1,82 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-01
+deciders: jochen
+reconstructed: false
+rests-on: 02-DECISIONS/0009-modules-and-the-graph.md
+---
+
+# 38. The mesh assigns the port, and a module does not care
+
+## The problem, as met
+
+A database module cannot start on a machine that runs the control plane. The mesh keeps its own
+store there and holds 5432; the module publishes 5432. Nothing notices until a container runtime
+three layers down says `port is already allocated`
+([`028`](../04-ISSUES/028-two-things-want-one-port-and-nothing-says-so/00-report.md)).
+
+A module cannot fix this by choosing better, because **a module cannot know what else is on the
+machine.** It is written once and assigned anywhere. Any number it picks is a guess about a
+machine it has never seen, and two modules guessing the same number is not a mistake either of
+them made.
+
+## The number is written three times, and nothing makes them agree
+
+Every module says its port in three places:
+
+| where | for | example |
+|---|---|---|
+| `listens` | the rule set that lets traffic in | `{port: 5432, from: mesh}` |
+| `serves` | what a consumer must know to connect | `{port: 5432}` |
+| a container's `ports` | what the runtime publishes | `"5432:5432"` |
+
+They agree today because one person wrote all three. Nothing checks it. A module whose `serves`
+said 5432 and whose container published 5433 would resolve, compose, apply, and hand every
+consumer a port that answers nothing.
+
+## The decision
+
+**The mesh assigns the machine-side port, and the module says only what it needs.** A module
+declares that a container port must be reachable and what it is for. Which number the machine uses
+is the mesh's to choose, because the mesh is the only thing that knows what else is there.
+
+**One source, and the other two are derived.** `serves` carries the assigned port so a consumer is
+told where to connect without the module having written it down; the rule set is computed from the
+same assignment. Three copies become one fact.
+
+**An assignment is made once and kept**, exactly as a credential is. A port that moved on every
+push would restart both ends each time and would hand consumers a number that was true when it was
+read.
+
+## Some ports cannot move, and that is a claim
+
+Mail is 25, submission is 587, IMAP over TLS is 993. A mail system on a strange port is not a mail
+system. So a module may say a port is **fixed by the protocol** rather than assigned.
+
+**A fixed port is exactly a claim** — the thing the mesh already has for what is singular on a
+machine: one seat, one display server, one artifact store. Two modules wanting 25 on one machine is
+the same shape as two wanting the seat, and gets the same answer: the second is refused, by name,
+when it is assigned rather than when it is applied.
+
+That is why this does not need a new mechanism so much as it needs the existing one pointed at
+ports.
+
+## What follows
+
+- **A module becomes portable in a way it was not.** Two databases on one machine stop being a
+  collision and become two assignments.
+- **The substrate has to be visible.** The mesh cannot assign around its own store while it has
+  never heard of it. What the bundle holds must be written down somewhere the assignment can read
+  — which the bundle does not say today.
+- **A refusal can be useful.** *25 is held by the mail system on this machine* is a sentence a
+  person can act on. `port is already allocated` is not.
+- **`serves` stops being written by hand**, which is a small vocabulary change with a large
+  consequence: what a consumer is told is now derived from what actually happened.
+
+## What this does not settle
+
+**Whether a module should publish to the machine at all.** Assignment makes publishing safe; it
+does not make it necessary. Consumers could instead reach a provider on the module's own network by
+name, with nothing published — which would make the question moot for anything inside the mesh, and
+would still leave it for anything reached from outside.

02-DECISIONS/0039-what-the-sdk-holds-and-refuses.md

@@ -0,0 +1,106 @@
+---
+topic: building it
+status: accepted
+date: 2026-09-03
+deciders: jochen
+reconstructed: false
+---
+
+# 39. What the SDK holds, and what it refuses
+
+_Reconciliation note (2026-09-05): supersedes the earlier "repository structure" decision, which the consolidation folded; no standalone record remains to point at, so body references to it now point at the nearest surviving record, [ADR 0015](0015-applications-live-in-their-own-repository.md)._
+
+## Context
+
+The earlier "repository structure" decision (folded in consolidation; see the reconciliation note
+above, and [ADR 0015](0015-applications-live-in-their-own-repository.md) as the nearest survivor)
+named `mesh-sdk` "contracts shared across tiers:
+types, not behaviour." That line is superseded here, because it draws the boundary in the wrong
+place. The boundary that matters is not *types versus behaviour* — it is **how often the thing
+changes**.
+
+The current SDK is the cautionary tale, and its failure is precise. `hal/sdk` holds all the
+code, including a per-module API client for every service (`clients/plex.ts`, `clients/gitea.ts`,
+…) and a per-module tool implementation for each (`tools/plex.ts`, …). Every module depends on
+the SDK, so **every edit to any of that per-module code rebuilds every module** — the cascade.
+The SDK is under constant maintenance precisely because it became the place all the volatile
+per-module logic accumulated.
+
+The root cause is worth stating exactly, because the fix follows from it: the pressure was never
+to share a client *between* modules. It was to share a client between one module's *own features*
+— plex's tools, its health check and its hooks all wanted the same `PlexClient` — and the only
+place to share code across a module's features was the global SDK. So **intra-module sharing
+leaked out as inter-module coupling.**
+
+## Decision
+
+The SDK holds the **stable spine** that modules build against, and earns its place by rarely
+changing. The test for membership is change-frequency, not kind.
+
+### What it holds
+
+- The **tool-serving harness** — the worker and registration mechanism, and the tool-definition
+  type. *How* a tool is declared and served is settled; it does not change when an individual
+  tool does.
+- The **messaging and event framework** — the broker client, the event consumer, the envelope.
+- The **contracts** — the manifest, declaration, provision and link shapes.
+- **Core primitives** — sealing and crypto, semver, the shared resolution helpers.
+
+These change rarely and deliberately. When one of them does change, a rebuild of everything is
+the *correct* outcome, because the contract every module shares has genuinely changed.
+
+### What it must not hold — the more important half
+
+- **A module's API client.** A Plex client, a Gitea client, a MinIO client belong in their
+  module. They change when that service's API or the module's use of it changes, which is often,
+  and which has nothing to do with any other module.
+- **A module's tool implementations.** Same reason, same place: in the module.
+- **Anything volatile** — anything that changes when one service's features change.
+
+The rule, stated so it can be applied without re-deriving it:
+
+> If editing a thing recompiles unrelated modules **and** it changes often, it does not belong
+> in the SDK.
+
+Both conditions are load-bearing. A rare change that cascades is fine — that is a contract, and
+the cascade is correct. A frequent change that stays local is fine — that is a module minding its
+own business. Only **frequent *and* cascading** is the disease, and per-module clients and tools
+are its carriers.
+
+### Where per-module shared code lives instead
+
+Code shared among a module's *own* features lives **in the module**. The default is the plainest
+thing that works: an ordinary shared file the features import — `plex/client.ts`, imported by
+`plex/tools/`. Within one module, features are files importing sibling files; no package
+boundary, no ceremony.
+
+A **module-local SDK** (a sub-package with its own version) is warranted only for the few modules
+whose shared surface is large enough to version on its own. It is the exception, not the shape.
+
+Either form gives the property the global SDK could not: editing a module's shared code rebuilds
+**that module and nothing else**.
+
+## Consequences
+
+- The cascade becomes **structurally impossible for module logic**. There is no longer an edge
+  from one module's internals to another, so the only thing that can rebuild everything is a real
+  change to a shared contract in the SDK — which is rare, and when it happens, is right.
+- The SDK is small and stable **by construction**, not by discipline. Its size is no longer a
+  thing anyone has to police.
+- **Converting a module from the current system is partly a de-coupling, not just a move.** Its
+  client and its tools are pulled *out* of the shared SDK and *into* the module. A conversion
+  that copied `clients/plex.ts` into the SDK's replacement would rebuild the exact mistake.
+- The host still does not import the SDK. It depends on nothing
+  ([ADR 0005](0005-the-node-host.md)) and **mirrors** the contracts rather than
+  importing them, exactly as its apply-shapes table already does deliberately. The SDK is shared
+  by the tiers that *can* share code; the host is not one of them.
+
+## References
+
+- The earlier "repository structure" decision — named the repositories; its `mesh-sdk`
+  description ("types, not behaviour") is superseded by this record (folded in consolidation;
+  nearest survivor [ADR 0015](0015-applications-live-in-their-own-repository.md)).
+- [ADR 0001](0001-mesh-brokers-nodes-host-agents-think.md) — the decomposition this serves: code
+  belongs to the boundary that owns it.
+- [ADR 0005](0005-the-node-host.md) — why the host mirrors the contracts instead of
+  importing the SDK.

02-DECISIONS/0040-what-a-module-is.md

@@ -0,0 +1,101 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-03
+deciders: jochen
+reconstructed: false
+extends: 0009-modules-and-the-graph.md
+---
+
+# 40. What a module is
+
+_Reconciliation note (2026-09-05): supersedes the earlier "grouped by domain" decision, which the consolidation folded into how-we-build.md; no standalone record remains to point at._
+
+## Context
+
+[ADR 0009](0009-modules-and-the-graph.md) settled that everything is a module, but never said what a
+module *is* beyond "a directory the mesh processes." That gap let the catalogue's breadth read as a
+smell: a module can carry a container, a built image, tools, a provisioner, migrations, health,
+config, seat claims, requires and provides — so much that the unit seemed ill-defined.
+The earlier "grouped by domain" decision (folded in consolidation; see the reconciliation note
+above) tried to organise modules by domain, which is the wrong axis. This record states what a module is, drawn from the cases that
+stress-tested it: the shell, i3-vs-sway, umami, and "database."
+
+## Decision
+
+**A module is one self-contained piece of software the mesh installs and manages** — everything
+needed to make that one thing real and integrable: what runs, the seats it claims, what it provides
+to other modules, what it requires from them, and what operates it.
+
+The **software is the module's identity.** Capabilities, seats and provisioned resources are the
+**relationships *between* modules**, not what a module is — and that is what binds a module into one
+thing. umami is bound by *being umami*: its container runs umami, its provisioner creates umami sites,
+its tools query umami, its `requires` gets umami a database. Every feature serves the one software.
+
+### The three relationships
+
+1. **Shared seat** — several modules fulfil a capability and coexist; one may be default. bash, zsh
+   and fish all join `shell`.
+2. **Exclusive seat** — modules contend for a single slot; one holds it. i3 (needs x11) and sway
+   (needs wayland) contend for `display-session`.
+3. **Provide / require** — a provider ships the **provisioner** that creates instances of the
+   resource it offers and returns sealed credentials; a consumer requires it and the mesh wires the
+   credential in. Symmetric: umami requires a database *and* provides analytics.
+
+### Interfaces are mesh-owned; providers adapt to them
+
+The mesh **defines the interface** for a capability — the provider-neutral contract of what a
+consumer receives and how it integrates. Both sides conform: a provider's provisioner **adapts** its
+software's real API to the mesh contract; a consumer depends on the **interface**, never on a
+provider. Swap one provider for another and the consumer does not change.
+
+### The naming rule — draw the interface at the consumer's real coupling
+
+Name a `provides`/`requires` at the **widest boundary across which the consumer genuinely does not
+care which implementation serves it**:
+
+- Where the consumer's coupling is thin — an analytics embed snippet and dashboard, opaque to it —
+  the mesh defines a neutral interface (`analytics`) and providers (umami, amumi) adapt. Swappable
+  across vendors.
+- Where the consumer **speaks a protocol** — a database's wire protocol and query dialect — the
+  interface *is* the protocol: `postgres-database`, `mssql-database`, `mongodb-database`. Swappable
+  only among protocol-compatible implementations, **never across**, because the application cannot
+  cross it either. "database" is not a capability; the protocol is.
+- **Never false genericity.** A name must not promise a swap the contract cannot deliver
+  ([research 005](../01-RESEARCH/005-domain-grouping/analysis.md)).
+
+This is [ADR 0027](0027-a-provision-names-what-the-consumer-is-coupled-to.md)'s rule made general — "names the protocol, not
+the product; a database names the engine because the app targets it" — with the reason stated: the
+contract sits where the coupling is.
+
+### What is not a module
+
+- A **library** (built against, never deployed — [ADR 0039](0039-what-the-sdk-holds-and-refuses.md)).
+- A **control-plane context** (the mesh itself — [ADR 0001](0001-mesh-brokers-nodes-host-agents-think.md)).
+
+A **swappable machine mechanism** (a firewall — ufw, nftables) *is* a module implementing a
+capability. The host hardcodes no firewall, supervisor, package manager or runtime; it owns only the
+generic apply primitives and platform detection, so it runs where none of those exist — an Android
+phone has no ufw, systemd, pacman or Docker.
+
+## Consequences
+
+- **Supersedes the earlier "grouped by domain" decision** (folded in consolidation; see the
+  reconciliation note above). Modules are
+  organised by their relationships (seats, provisions), not grouped into domain folders.
+- **Refines [ADR 0009](0009-modules-and-the-graph.md).** Everything the mesh runs and integrates is
+  a module — but a module is defined by the *software it delivers*, not by being a bucket of features.
+- The target is a **self-fulfilling mesh**: declared wants bound to swappable modules, provisioners
+  wiring credentials, nothing hardcoded. The control plane's whole job is the binding.
+- Converting a module from the old system includes pulling its per-module code out of the shared SDK
+  ([ADR 0039](0039-what-the-sdk-holds-and-refuses.md)) and shipping its provisioner as an adapter to a
+  mesh interface — a de-coupling, not just a move.
+
+## References
+
+- [ADR 0009](0009-modules-and-the-graph.md) — everything is a module; this says what one is.
+- [ADR 0001](0001-mesh-brokers-nodes-host-agents-think.md) — contexts are the mesh, not modules.
+- The earlier "grouped by domain" decision — superseded (folded in consolidation; see the note above).
+- [ADR 0027](0027-a-provision-names-what-the-consumer-is-coupled-to.md) — protocol-not-product, generalised here.
+- [ADR 0039](0039-what-the-sdk-holds-and-refuses.md) — per-module code lives in the module.
+- [research 011](../01-RESEARCH/011-the-module-graph/00-overview.md) — the graph of these relationships.

02-DECISIONS/0041-events-are-a-relationship.md

@@ -0,0 +1,101 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-03
+deciders: jochen
+reconstructed: false
+extends: 0040-what-a-module-is.md
+---
+
+# 41. Events are a relationship, the lighter sibling of provisioning
+
+## Context
+
+[ADR 0040](0040-what-a-module-is.md) names two relationships between modules — seats and
+provide/require (provisioning). A third is latent in the mesh and worth making first-class: the
+broker every node already runs ([ADR 0002](0002-nodes-communicate-over-a-broker.md)) can carry a
+module's activity as **events**, which any other module reacts to. A logger that writes an audit
+trail, a module that acts when another module acts, observability — all of it is one mechanism, and
+today it is ambient rather than declared.
+
+## Decision
+
+**A module emits events and consumes events, and both are declared** — parallel to `provides` /
+`requires`, so the mesh knows the event graph the same way it knows the provisioning graph.
+
+### Events are provisioning's lighter sibling
+
+| | provisioning | events |
+|---|---|---|
+| shape | **1:1**, a provider creates a resource *for* one consumer | **1:many**, a module emits, any number listen |
+| credential | yes — sealed, per consumer | none — it is broadcast |
+| machinery | a provisioner (the reconcile adapter) | nothing but the broker's topic routing |
+| declared as | `provides` / `requires` | `emits` / `consumes` |
+
+Because an event is broadcast and credential-free, there is no provisioner and no per-consumer
+setup — only a subscription. That is why it is the *lighter* relationship, and why most
+inter-module reaction should be an event, not a provision.
+
+### An event carries what an audit needs
+
+Every event carries its **type** (a dotted topic key, so listeners match by prefix), its **source**
+module, the **node** it came from, and the **time**. A body follows. The metadata is not optional:
+a reaction may only need the body, but an audit trail needs to know who did what, where and when,
+and an event that cannot answer that is not auditable.
+
+### The audit logger is just a consumer of everything
+
+A logger that records the whole mesh's activity is **not a privileged component** — it is an
+ordinary module that consumes `#` (every event) and writes them down. It holds no special access;
+it only listens widely. That it falls out of the model with no new machinery is the check that the
+model is right.
+
+### `consumes` is validated like `requires`
+
+A `consumes` for an event that **nothing** `emits` is a dangling edge, and the mesh refuses it
+before deploy — the same rule that catches a `requires` for a resource nothing provides
+([research 011](../01-RESEARCH/011-the-module-graph/00-overview.md)). A listener waiting for an
+event that can never arrive is a silent failure, and this repository's whole discipline is against
+silent failure.
+
+### One runtime serves all three
+
+The per-node module runtime that serves a module's tools also wires its `consumes` (subscribe,
+dispatch to the handler) and lets its code `emit`. Tools are *invoked* (request/reply), resources
+are *provisioned* (1:1, credentialed), events are *emitted and consumed* (1:many, broadcast) —
+three relationships, one broker, one runtime, all declared on the manifest.
+
+## Consequences
+
+- The mesh gains a declared **event graph** alongside the provisioning graph — visible, validated,
+  reasoned over.
+- **Reaction becomes the default coordination**: a module acts on another's event without either
+  knowing the other, and without a credentialed link. Coupling drops.
+- An **audit trail** is a module, not a platform feature — and can be swapped, extended or run more
+  than once (a file logger and a queryable one) with no change to anything that emits.
+- The runtime must dispatch a module's event handlers as well as its tools; that generalisation is
+  small (both arrive by importing the module's entrypoint) but it is real work.
+
+## Progressive insight
+
+> **Progressive insight — 2026-09-26.** *"No provisioner and no per-consumer setup" was a fact
+> about the transport, and the transport changed.* This record's table says an event's machinery is
+> "nothing but the broker's topic routing", and the text that an event needs "no per-consumer setup
+> — only a subscription". That was true of a topic exchange, where a binding cost nothing and the
+> broker fanned out. On NATS
+> ([ADR 0106](0106-the-bus-is-nats.md)) a subscription is a **durable consumer**: a real object
+> with a name, an ack policy, a delivery limit and its own ack subject, created when a module is
+> assigned and removed when it is not. Per-consumer setup exists, and the controller does it.
+>
+> The decision is untouched — events are declared on both sides, 1:many, credential-free, and
+> still provisioning's lighter sibling; the lightness is now relative rather than absolute.
+> [ADR 0126](0126-a-module-declares-its-own-seats.md) adds the relationship this record's two
+> columns had no room for: work addressed to a role, where exactly one holder must act.
+
+## References
+
+- [ADR 0002](0002-nodes-communicate-over-a-broker.md) — the broker events ride.
+- [ADR 0040](0040-what-a-module-is.md) — the relationships this extends.
+- [ADR 0039](0039-what-the-sdk-holds-and-refuses.md) — `emit`/`on` are stable sdk surface; the
+  broker binding and the runtime are not.
+- [research 011](../01-RESEARCH/011-the-module-graph/00-overview.md) — the graph these edges join.

02-DECISIONS/0042-the-shape-of-an-event-on-the-wire.md

@@ -0,0 +1,116 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-03
+deciders: jochen
+reconstructed: false
+extends: 0041-events-are-a-relationship.md
+---
+
+# 42. The shape of an event on the wire
+
+## Context
+
+[ADR 0041](0041-events-are-a-relationship.md) made events a relationship — `emits`/`consumes`, the
+graph, the audit logger. It did not say what an event *is* on the broker: the exchanges, the
+routing keys, the headers, the queues and their configuration. That shape is a contract every
+emitter and consumer conforms to, exactly as [ADR 0010](0010-delivery.md)
+is for declarations — and it was being decided ad-hoc in code. This settles it, so the sdk and the
+runtime implement one contract and a module never reinvents it.
+
+## Decision
+
+### Two exchanges, kept apart
+
+- **`mesh.events`** — a durable topic exchange. Every event rides it: module, mesh and node.
+- **`mesh.rpc`** — a durable topic exchange. Tool invocations (request/reply) ride it.
+
+Kept separate because RPC is not an event: a `#` subscription on `mesh.events` is then a complete
+audit of what happened, with none of the invocation traffic.
+
+### The routing key is the event type, namespaced by origin
+
+Dotted and hierarchical — `<origin>.<name>.<event…>` — with three reserved origins:
+
+- `module.<module>.<event>` — `module.umami.site.created`
+- `mesh.<context>.<event>` — `mesh.delivery.deployed`, `mesh.provisioning.granted`
+- `node.<node>.<event>` — `node.anchor.joined`, `node.anchor.unreachable`
+
+Topic matching gives a consumer `node.*.joined`, `module.umami.#`, or `#`. The origin roots are
+reserved; everything after is the emitter's own namespace.
+
+### Metadata in headers, payload in the body
+
+An event's identity and provenance are AMQP **headers**, so a consumer — or the broker, or an
+audit tool — reads who/when/what without parsing the body, and the body is only the domain payload.
+
+**Required headers**
+
+| header | meaning |
+|---|---|
+| `x-event-id` | a unique id — for dedup and audit (delivery is at-least-once, below) |
+| `x-source` | the emitter: the module, context or node name |
+| `x-node` | the node it was emitted from |
+| `x-time` | emit time, RFC-3339 |
+| `content-type` | `application/json` |
+
+**Optional headers**
+
+| header | meaning |
+|---|---|
+| `x-causation-id` | the event or command that caused this one — tracing |
+| `x-schema` | a version of the body's shape, so a body evolves without silent misreads |
+
+The routing key already carries the type; it is not duplicated as a header. An **unknown `x-`
+header is ignored, not refused** — unlike a declaration, an event is observed by parties that need
+not all understand every header, and refusing would couple every consumer to every emitter's
+additions.
+
+### Messages are persistent
+
+Events are published persistent (delivery-mode 2). An audit trail that loses events on a broker
+restart is not one, and the cost is disk the broker already spends on everything durable.
+
+### Queues: one per consumer, durable, dead-lettered
+
+- **A consumer's queue** is `<node>.<module>.events`, durable, bound to that module's consumed
+  patterns. Durable so a restart does not drop what arrived while it was down. **Manual ack** after
+  the handler succeeds — at-least-once.
+- **Prefetch** bounds in-flight work (default 32) so one slow consumer does not pull the whole
+  backlog into memory.
+- **A dead-letter exchange** `mesh.events.dead` receives a message rejected past a redelivery limit,
+  so a poison event is set aside for inspection rather than looping forever or vanishing silently.
+- **The audit logger's queue** `<node>.audit-logger.events`, bound to `#`, is the same shape —
+  durable, persistent, dead-lettered — because completeness is its whole job.
+- **RPC reply queues** are exclusive, auto-delete and server-named; **RPC serve queues**
+  `serve.<key>` are durable and shared, so several runtimes serving one tool key compete rather than
+  each answer.
+
+### At-least-once, and consumers are idempotent
+
+A handler may see an event twice — a redelivery after a crash between doing the work and acking.
+Consumers must be idempotent, and `x-event-id` is what makes dedup possible. **Exactly-once is not
+offered**: it is a promise no broker keeps honestly, and saying so is better than pretending.
+
+## Consequences
+
+- The event shape is a versioned, enforced contract, not conventions each module reinvents. The
+  sdk's `emit`/`on` and the runtime's AMQP binding implement it; a module never sees an exchange or
+  queue name.
+- Metadata-in-headers means the body is exactly the domain payload, and a consumer that only wants
+  provenance never parses it.
+- Adding a header or an origin root widens the contract and is reviewed as one — the discipline
+  [ADR 0010](0010-delivery.md) applies to the
+  declaration vocabulary.
+- The sdk's first cut carried source/node/time in the *body*; this supersedes that — they move to
+  headers. That is code to align, in `mesh-sdk` (`emit`/`on`) and `mesh-tools` (the binding, queue
+  config, dead-letter).
+
+## References
+
+- [ADR 0041](0041-events-are-a-relationship.md) — events as a relationship; this is their wire shape.
+- [ADR 0010](0010-delivery.md) — the precedent: a wire
+  contract, versioned, additions reviewed as security.
+- [ADR 0002](0002-nodes-communicate-over-a-broker.md) — the broker.
+- [ADR 0039](0039-what-the-sdk-holds-and-refuses.md) — `emit`/`on` are stable sdk surface; the
+  binding, queue config and dead-letter are the runtime's, not the sdk's.

02-DECISIONS/0043-a-module-broker-account-is-scoped-by-emits-and-consumes.md

@@ -0,0 +1,101 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-04
+deciders: jochen
+reconstructed: false
+extends: 0041-events-are-a-relationship.md
+---
+
+# 43. A module's broker account is scoped by what it emits and consumes
+
+## Context
+
+[ADR 0041](0041-events-are-a-relationship.md) made events a relationship — `emits` and `consumes`
+on the manifest. [ADR 0042](0042-the-shape-of-an-event-on-the-wire.md) gave them a wire shape — the
+`mesh.events` exchange, the durable per-consumer queue, the reserved routing-key origins. Neither
+said how a module *reaches* the broker: what account it holds, and what that account is allowed to
+do.
+
+As the code stands, there is no answer. The mesh can provision a **node** account (at enrolment)
+and a **builder** account (scoped to the build queue), and it can *deliver* any module a sealed
+own-secret at a declared path — but it has no way to provision a broker **account** for a general
+module. A module that declares `own-secrets: {broker: …}` and nothing more receives thirty-two
+random bytes, not a credential. So on the broker, `emits` and `consumes` are enforced by nothing: a
+running module could bind any queue, consume any pattern, and publish under any origin, and the
+manifest that says otherwise would be describing a boundary no code draws — the exact shape of fault
+[04-ISSUES/003](../04-ISSUES/003-firewall-scope-is-read-by-no-code/00-report.md) records, a scope
+declared in manifests and read by nothing.
+
+This settles it, so a module's place on the bus is a thing the broker enforces rather than a thing
+the manifest merely claims.
+
+## Decision
+
+### A module gets a broker account when it is assigned, and its permissions are the manifest
+
+When the mesh assigns a module to a node it provisions a broker account for that module on that node,
+sealed to the node ([ADR 0004](0004-a-node-and-how-it-joins.md)) and delivered as the
+module's `own-secrets` broker — `amqps://` with the mesh's fingerprint, the shape
+[ADR 0042](0042-the-shape-of-an-event-on-the-wire.md) already carries. The account's permissions are
+derived from the manifest, and are exactly these:
+
+- **What it consumes.** Read on `mesh.events`, and configure-and-read on its own queue
+  `<node>.<module>.events` bound to the patterns in `consumes`. It cannot bind or read another
+  module's queue. A module that consumes nothing gets no read on the events exchange at all.
+- **What it emits.** Write to `mesh.events`, restricted to routing keys under its own origin,
+  `module.<name>.*`. It cannot publish as another module, and cannot publish under the reserved
+  `mesh.*` or `node.*` origins — those belong to the mesh and the host (ADR 0042). A module that
+  emits nothing gets no write.
+- **Nothing else.** The events account reaches `mesh.events` and that module's own queue, and no
+  more. Tool serving and calling over `mesh.rpc` is a separate grant on the same principle — a
+  module serves the tool keys it declares and calls the ones it is bound to — and is scoped the same
+  way rather than folded in here.
+
+### Consuming everything is a privilege, granted deliberately
+
+`consumes: ["#"]` — the audit logger — is read across the whole bus: every module's events, the
+mesh's, every node's. That is not a pattern like any other; it is the power to see everything, and
+the account is where it becomes visible. The grant that lets one module read the entire bus is one
+the mesh issues on purpose and can be audited — the answer to *who can read everything* is a row, not
+a guess — rather than a breadth any manifest acquires by typing a single character. A `#` consume is
+a reviewed grant, not a default one.
+
+### The account is how the declaration is enforced
+
+Because the account can do only what `emits` and `consumes` name, the broker itself refuses a module
+that tries to consume a queue it did not declare or emit under an origin it does not own. That is what
+makes an event relationship a rule and not a comment — the discipline that a stated rule says how it
+is checked. A manifest that over-declares grants more than the module uses, which is visible and
+reviewable; one that under-declares makes the module fail closed at the broker, which is the safe
+direction to be wrong in.
+
+## Consequences
+
+- The control plane gains a **generic module broker-account**, derived from the manifest. The
+  builder stops being a special case: its access to the build queue becomes an ordinary expression of
+  what it consumes and serves, not a bespoke account method. One rule, and the builder is an instance
+  of it.
+- The runtime reads its credential from a file (the broker own-secret), `amqps://` verified against
+  the mesh's fingerprint. The `guest` account is for raising the substrate, never for a module — a
+  module documented as holding its own credential and handed the broker's administrative one is worse
+  than one with no credential story at all.
+- `emits` and `consumes` stop being advisory. They are the module's authority on the bus, so the
+  manifest is now a security boundary and is reviewed as one, the discipline
+  [ADR 0010](0010-delivery.md) applies to the declaration
+  vocabulary.
+- *Who can read the whole bus* becomes an answerable question, because `#` is a grant and not an
+  accident.
+
+## References
+
+- [ADR 0041](0041-events-are-a-relationship.md) — events are a relationship; this scopes the account
+  by that relationship.
+- [ADR 0042](0042-the-shape-of-an-event-on-the-wire.md) — the wire this account secures: the queue,
+  the origins, the `amqps` credential shape.
+- [ADR 0004](0004-a-node-and-how-it-joins.md) — the link is the security boundary; a
+  module's account is sealed to its node the same way a node's is.
+- [ADR 0010](0010-delivery.md) — a declaration is owned
+  and its additions reviewed; a module's broker permissions are that discipline applied to the bus.
+- [04-ISSUES/003](../04-ISSUES/003-firewall-scope-is-read-by-no-code/00-report.md) — a scope declared
+  in manifests and enforced by no code: the fault this decision closes for events.

02-DECISIONS/0044-a-public-name-is-provisioned-like-any-capability.md

@@ -0,0 +1,94 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-04
+deciders: jochen
+reconstructed: false
+extends: 0027-a-provision-names-what-the-consumer-is-coupled-to.md
+---
+
+# 44. A public name is provisioned, not registered by hand
+
+## Context
+
+The mesh names and resolves its own machines internally: the overlay generates
+`<service>.<node>.<suffix>` wildcards, dnsmasq answers them (`wildcard-resolution`), and the mesh
+issues a certificate for each internal name. A service reachable at a *public* domain —
+`plex.example.com`, not `plex.anchor.internal` — needs three things that machinery does not give it:
+
+- a **public DNS record** at a registrar or DNS provider, so the name resolves on the internet;
+- a **publicly-trusted certificate** for it, because the mesh's own authority is trusted by nobody
+  outside the mesh;
+- and routing from that name to the module — which the reverse proxy already does: a module
+  `requires` the `route` capability and the proxy provides it, routing by the host it was asked for.
+
+The routing exists. The public DNS record does not: the mesh has no way to make a name resolve on
+the public internet, so today that is a step someone does by hand at a DNS provider, outside the
+mesh, remembered nowhere. A public name is therefore the one part of reaching a service that the
+declaration graph cannot grant or withdraw — which means it is created once and outlives whatever it
+was for, the shape of drift this project exists to remove.
+
+## Decision
+
+### A public name is a capability, requested like any other
+
+A module reachable at a public host declares `requires: ["public-dns"]` and contributes the hostname
+it wants — beside `requires: ["route"]`, which exposes it through the proxy. The name is then
+provisioned on declaration ([ADR 0027](0027-a-provision-names-what-the-consumer-is-coupled-to.md)): created
+when the module is assigned, removed when it is withdrawn, reconciled like every provision.
+
+### The interface is neutral; the providers are the registrars
+
+`public-dns` is drawn at the consumer's coupling: the consumer wants *a public name that resolves to
+me*, and does not care whether Cloudflare, Route 53 or a registrar's own API puts the record there.
+So the interface is neutral and the providers are provider-scoped — `cloudflare-dns`,
+`route53-dns`, `porkbun-dns` — each implementing the one `public-dns` contract, the same way a
+neutral database coupling is answered by `postgres-database` and `mssql-database`. A module names
+`public-dns`; it never names a registrar.
+
+### The record points at the mesh's public ingress, not at the node
+
+What the name resolves to is the address the reverse proxy answers on, not the consuming machine's.
+A public service is reachable only *through* the proxy — the proxy holds the `route` grant and routes
+by host to the module — so the public name must resolve to the proxy. `public-dns` and `route` are
+the two halves of one public exposure: the name, and what the name reaches.
+
+### The record is a fact, not a secret
+
+A DNS record is public by definition, so the grant returns the fully-qualified name and its TTL and
+nothing sealed. The only secret is the provider's own API credential, which is the provider module's
+own-secret and never leaves it — the module that wanted the name never sees it.
+
+### Events
+
+The provider emits `module.<provider>.record.created` and `module.<provider>.record.removed`
+([ADR 0041](0041-events-are-a-relationship.md)), so *which names the mesh publishes, and where* is a
+question answered from the event trail and the grants, not from a folder of records edited at a
+provider.
+
+### The public certificate is the proxy's, and is named here only to pair it
+
+A public name without a publicly-trusted certificate is reachable and not trusted — the same pairing
+the internal name and the mesh-issued certificate already have. Obtaining that certificate (ACME
+against the now-resolving public name) is the reverse proxy's to do, and its mechanism is its own
+decision; it is named here so the pairing is not forgotten, not resolved here.
+
+## Consequences
+
+- A public name is created and torn down with the module, so it cannot outlive it, and the mesh can
+  say which public names it publishes without anyone reading a registrar's dashboard.
+- Adding a registrar is adding a provider that answers `public-dns`; the modules that want names do
+  not change.
+- Public exposure of a service is a trio of separate, declared, enforced relationships: the firewall
+  opens the proxy's public port ([ADR 0045](0045-a-machine-firewall-is-the-sum-of-what-it-listens-on.md)),
+  `route` routes the host to the module, and `public-dns` makes the host resolve.
+
+## References
+
+- [ADR 0027](0027-a-provision-names-what-the-consumer-is-coupled-to.md) — a capability is provisioned on
+  declaration; a public name is one.
+- [ADR 0045](0045-a-machine-firewall-is-the-sum-of-what-it-listens-on.md) — the firewall, the other
+  half of the reachability question this was asked with.
+- [ADR 0041](0041-events-are-a-relationship.md) — the provider's record events.
+- [ADR 0040](0040-what-a-module-is.md) — a provider and its interface; the neutral-interface,
+  scoped-provider naming this follows.

02-DECISIONS/0045-a-machine-firewall-is-the-sum-of-what-it-listens-on.md

@@ -0,0 +1,93 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-04
+deciders: jochen
+reconstructed: false
+extends: 0005-the-node-host.md
+---
+
+# 45. A machine's firewall is the sum of what its modules listen on
+
+## Context
+
+The reverse proxy is a *provider*: a module `requires` the `route` capability and a running proxy
+provides it, routing traffic by name and reaching back to the consumer. A fair question follows —
+is the firewall the same shape? Should a module *register* a port with a firewall provider the way
+it requests a route?
+
+It should not, and the difference is the point. A reverse proxy is a service another component
+performs; a firewall is a property of the machine — a packet filter the host applies to itself.
+Modelling it as a provider would invent a credential and a reach-back for something that has neither.
+
+And the mesh already has the registration: a module declares `listens: [{ port, from }]` — the port
+it accepts connections on, and from where. That *is* how a service says it wants a port open. What is
+missing is not a model but enforcement. [04-ISSUES/003](../04-ISSUES/003-firewall-scope-is-read-by-no-code/00-report.md)
+records that a `scope:` key five manifests carry is read by no code: a manifest can appear to
+restrict a port and restrict nothing — the exact fault
+[how-we-build.md](../00-META/how-we-build.md) names, *an unenforced rule is indistinguishable from a
+wrong one*, made worse because the declaration reads as a restriction.
+
+## Decision
+
+### The firewall is derived and host-applied, not a provider
+
+A machine's firewall is the sum of what the modules assigned to it declare they listen on, computed
+by the host and applied as one of its owned resources ([ADR 0005](0005-the-node-host.md):
+the host applies, it does not decide; [ADR 0010](0010-delivery.md):
+the declaration is owned resources). It is not a capability, not a per-consumer grant — opening a
+port is a declarative fact about a machine, so it is computed and applied, not requested and
+credentialed.
+
+### `from` is the whole of public-versus-internal
+
+The distinction the question is really about lives in `from`:
+
+- `listens: [{ port: 5432, from: mesh }]` — open to the private overlay only.
+- `listens: [{ port: 443, from: anywhere }]` — open to the public internet.
+
+A module registers a port on the firewall by listening on it and saying from where. There is no
+separate firewall capability, because the firewall is not a thing that reaches back or holds a
+secret; it is the machine's own filter over the ports its modules named.
+
+### The host enforces it both ways, and unknown keys are refused
+
+A port a module listens on is opened to exactly the scope it named; a port nothing declares is
+closed. And a key the firewall does not read — the `scope:` of issue 003 — is refused at the
+manifest, not accepted and ignored, so a declaration that reads as a restriction is one. This is the
+discipline [ADR 0043](0043-a-module-broker-account-is-scoped-by-emits-and-consumes.md) applied to the
+broker account, applied here to the packet filter: the declaration is the enforcement, or it is a
+comment.
+
+### A public service is exposed through the proxy, not by opening its own port
+
+Reaching the public internet is normally not `from: anywhere` on the service's own port. The service
+listens `from: mesh` — only the proxy reaches it — and `requires: route`, so the sole machine with a
+public opening is the one running the reverse proxy, and the service is exposed by name through it.
+`from: anywhere` is the deliberate direct-exposure case, for a service that is its own front door.
+
+## Consequences
+
+- Issue 003 is closed: the firewall is computed from `listens` and enforced, so a declared scope is
+  real and an undeclared port is shut. Rejecting unknown manifest keys is the general fix, of which
+  the `scope:` key was one instance.
+- The firewall and the reverse proxy stop being confused for one model: the firewall is the machine's
+  filter (host-derived from `listens.from`); `route` is a name-router (a provider); the public DNS
+  name is a third thing ([ADR 0044](0044-a-public-name-is-provisioned-like-any-capability.md)). A
+  public service uses all three.
+- The modelling question is answered: a module registers a port by declaring `listens`, and reaches
+  the public internet by name through `route` + `public-dns` — never by the firewall being a
+  provider.
+
+## References
+
+- [ADR 0005](0005-the-node-host.md) — the host applies; the firewall is one of
+  the things it applies.
+- [ADR 0010](0010-delivery.md) — the firewall is a derived
+  owned resource, not a grant.
+- [ADR 0043](0043-a-module-broker-account-is-scoped-by-emits-and-consumes.md) — the same discipline:
+  a declaration is enforced, or it is a comment.
+- [ADR 0044](0044-a-public-name-is-provisioned-like-any-capability.md) — the public name, the other
+  half of the reachability question this was asked with.
+- [04-ISSUES/003](../04-ISSUES/003-firewall-scope-is-read-by-no-code/00-report.md) — the unenforced
+  `scope:` this closes.

02-DECISIONS/0046-a-module-configuration-is-its-assignments-not-its-manifest.md

@@ -0,0 +1,88 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-04
+deciders: jochen
+reconstructed: false
+extends: 0027-a-provision-names-what-the-consumer-is-coupled-to.md
+---
+
+# 46. A module's configuration is its assignment's, not its manifest's
+
+## Context
+
+A module is assigned to a node — `assign <node> <module>`, always to a machine; there is no
+assignment to the mesh. "Mesh" is a *scope*, not a place: a `provides` or a `claim` scoped `mesh`
+reaches the whole mesh, but the module still runs on a node. So the two kinds of thing a module can
+carry are the manifest (what the module *is*) and, separately, what it should do *here* — which
+differs by deployment and by node.
+
+The mesh already has the second: **settings**. `settings set <module> [--node <node>]` — with a node
+it is that machine's, without it the whole mesh's — layered over what the module declares and applied
+at resolution, changeable without editing the module and without a rebuild. That is the surface a
+meshboard would edit.
+
+But settings today reach only a module's **config-file content** (a mergeable file the module owns).
+Configuration that is not a file has been landing in the manifest instead, statically — a registrar's
+zone and domain, the address public names point at, and, most sharply, `listens.from`. That last one
+is the tell: whether a port is open to the private overlay or to the public internet is a
+*per-node deployment choice* — the same database internal on one machine and public on another — and
+a value fixed in the manifest is one value for every machine, so it cannot be. Static configuration in
+the manifest is configuration in the wrong place: it cannot vary per node, and it cannot change
+without a new module version.
+
+## Decision
+
+### The manifest is identity and defaults; the assignment's settings are the configuration
+
+A module's manifest declares what it is — what it provides, requires and claims, the shape of its
+resources — and, for anything configurable, a **default**. The values that make a running instance
+*this* instance are settings, carried by the assignment: per-node, or mesh-wide when no node is named,
+applied over the defaults at resolution. Change one and the next reconcile carries it; nothing is
+edited on a machine and nothing is rebuilt.
+
+### Settings drive the configurable fields the manifest marks, not only file content
+
+Settings extend beyond a config file's content to the manifest fields a module declares settable —
+foremost:
+
+- **`listens.from`**: a module declares its safe default (`from: mesh`), and a per-node setting
+  raises or lowers it. postgres declares `listens: [{ port: 5432, from: mesh }]`; on the machine that
+  should expose it, a setting makes that port `from: anywhere`. Same module, different exposure, and
+  the firewall ([ADR 0045](0045-a-machine-firewall-is-the-sum-of-what-it-listens-on.md)) is computed
+  from the effective value, so the packet filter follows the setting.
+- **A provider's own configuration**: a registrar's zone, domain and the ingress its names point at
+  ([ADR 0044](0044-a-public-name-is-provisioned-like-any-capability.md)) are mesh-wide settings, not
+  manifest constants — one mesh's Cloudflare zone is not another's, and the module description is the
+  same for both.
+
+### Unset is the default, and an unknown setting is refused
+
+A field with no setting keeps the manifest's default, so a module runs correctly configured by nobody.
+A setting that matches no settable field — like a config value that reaches no file today — is named,
+not silently dropped, so a misspelled setting is found rather than believed (the discipline of
+`UnusedSettings`, and of [ADR 0043](0043-a-module-broker-account-is-scoped-by-emits-and-consumes.md):
+a declaration is enforced or it is a comment).
+
+## Consequences
+
+- The postgres case works: one module, `from: mesh` by default, `from: anywhere` where a setting says
+  so — internal on ace, public on novox, changeable live.
+- Provider modules stop carrying a mesh's specifics: `cloudflare-dns` describes *a Cloudflare
+  registrar*, and *which* zone and ingress is a setting, so the same module serves every mesh.
+- Configuration becomes a thing a meshboard manages — set per node or mesh-wide, applied on the next
+  reconcile — rather than a manifest edit and a rebuild ([ADR 0011](0011-managed-files-are-generated-never-edited.md):
+  the way you change a managed thing is not by editing it).
+- What a manifest may not do is grow a value that differs per machine; if it differs per machine it is
+  a setting, and the manifest holds only the default.
+
+## References
+
+- [ADR 0027](0027-a-provision-names-what-the-consumer-is-coupled-to.md) — what is provisioned on
+  declaration; its per-instance values are the assignment's.
+- [ADR 0011](0011-managed-files-are-generated-never-edited.md) — a managed thing is changed through
+  the mesh, not by editing it; settings are that, for configuration.
+- [ADR 0045](0045-a-machine-firewall-is-the-sum-of-what-it-listens-on.md) — the firewall follows the
+  effective `listens.from`, so making `from` a setting makes exposure a setting.
+- [ADR 0044](0044-a-public-name-is-provisioned-like-any-capability.md) — the provider whose zone and
+  ingress are settings, not manifest constants.

02-DECISIONS/0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md

@@ -0,0 +1,87 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-04
+deciders: jochen
+reconstructed: false
+extends: 0043-a-module-broker-account-is-scoped-by-emits-and-consumes.md
+---
+
+# 47. A module runs its code as its own process, with its own account
+
+## Context
+
+A module is one self-contained thing ([ADR 0040](0040-what-a-module-is.md)), and it gets a broker
+account scoped to what it emits and consumes ([ADR 0043](0043-a-module-broker-account-is-scoped-by-emits-and-consumes.md)).
+The catalogue now gives modules **tools** and **events** — real code, in the module ([ADR 0039](0039-what-the-sdk-holds-and-refuses.md)) —
+but nothing has said what *runs* that code. The audit-logger showed one shape and was treated as an
+exception: a container running the tool runtime carrying the module's compiled code, holding the
+module's own scoped account. Every module with tools or events needs the same, and the tempting
+alternative does not work.
+
+**A node-wide runtime that loaded every assigned module's code cannot hold a per-module account.** It
+would run under one account with the union of every module's permissions — able to emit as any of
+them and read any of their queues — which is exactly the isolation ADR 0043 exists to draw. So the
+runtime is per-module, not per-node, and treating the audit-logger as special left the other
+modules' code with nothing to run it: the conversion produced tools and events that, as it stands,
+never execute.
+
+## Decision
+
+### A module with tools or events runs a process of its own
+
+A module that has tools or events runs a **runtime process** — a container, the tool runtime carrying
+that module's compiled code — assigned and started like the module it is, holding the single broker
+account the mesh scoped to it (ADR 0043). One module, one process, one account.
+
+### It serves its tools, each on its own key
+
+A tool is served on its own key (`serve.<tool>`), and a caller invokes a named tool. Only the module
+that serves it answers, and the module's account is scoped to exactly its tool keys — so one module
+cannot answer another's calls, the isolation ADR 0043 gives events extended to tools. This supersedes
+a single `tools.invoke` endpoint that dispatched by name: that shape assumed one runtime for the
+whole node, and per-module runtimes competing on one key would each be handed calls for tools they do
+not have.
+
+### It runs its events in the same process, under the same account
+
+Emitting under the module's own origin and consuming its own queue ([ADR 0042](0042-the-shape-of-an-event-on-the-wire.md))
+happen in that same process, with that same account — not a second one to scope and seal. A module's
+tool code, its event code and, for a provider, its provisioner are the one module's code and run as
+the one module's process.
+
+### The runtime image is the tool runtime plus the module's code
+
+Built from the module's source like any module image — the audit-logger's shape, made the rule, not
+the exception. The module declares a `container` for it carrying `MESH_BROKER_FILE` (its sealed
+credential, ADR 0043) and its compiled code. A module with **neither** tools nor events runs no such
+process: a plain service module — the plex *server*, dnsmasq the resolver — is its service and files
+and nothing more. A module that is both a service and code declares both containers: the service, and
+the runtime beside it.
+
+## Consequences
+
+- The catalogue's tools and events become runnable: each tools-or-events module gains a runtime
+  container with its scoped credential, and the audit-logger stops being special. Until this, the
+  converted modules held code with nothing to execute it.
+- A process, and a small image, per tools-or-events module. That is the cost of ADR 0043's isolation:
+  one account per module means one process per module. It is paid deliberately — a shared runtime is
+  cheaper and cannot be scoped, and a mesh where any module can emit as any other is not one worth the
+  saving.
+- `serve.<tool>` per key replaces the single `tools.invoke` dispatch. The sdk's serving and a module's
+  account scope both come to name tools individually.
+- **A provider's provisioner is a runtime process too.** It already runs as its own container; its
+  events (`bucket.created`, `database.provisioned`) belong to *that* process and need the same
+  credential. So a provisioner that emits carries `MESH_BROKER_FILE` and its scoped account like any
+  runtime — or it does not emit. (This is the fix for provisioners that emit today with no broker
+  bound: the emit is a runtime's, and the provisioner is a runtime.)
+
+## References
+
+- [ADR 0040](0040-what-a-module-is.md) — a module is one self-contained thing; its code runs as one
+  process.
+- [ADR 0043](0043-a-module-broker-account-is-scoped-by-emits-and-consumes.md) — the scoped account
+  this process holds, and the isolation that makes it per-module.
+- [ADR 0042](0042-the-shape-of-an-event-on-the-wire.md) — the events this process runs, and the
+  `serve.<key>` queue tools now use.
+- [ADR 0039](0039-what-the-sdk-holds-and-refuses.md) — the code lives in the module; this runs it.

02-DECISIONS/0048-a-provider-creates-the-credential-the-mesh-minted.md

@@ -0,0 +1,131 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-05
+deciders: jochen
+reconstructed: false
+---
+
+# 48. A provider creates the credential the mesh minted, and seals nothing
+
+## Context
+
+A provider module stands up a per-consumer resource — a database, a cache bucket, an object
+store user — and the consumer must end up holding a credential that authenticates against it.
+Building the module runtime (ADR 0047: a module runs its own code as its own process under its
+own account), the provider's provisioner was run for the first time as a delivered thing, and
+it did not work. It reads a seal key from the environment that nothing sets, and it seals every
+credential it produces to that key with a symmetric passphrase.
+
+Tracing the credential's path turned up something larger than a missing key. **The provisioner
+harness the whole catalogue is built on describes a credential flow the mesh does not have, and
+duplicates — incorrectly — one it does.**
+
+What the sdk's `runProvisioner` does today:
+
+- reads request files named `*.grant.json` — which nothing in the mesh writes;
+- calls an adapter whose `create` **generates its own password** and returns it;
+- seals that password with a symmetric key (`$MESH_SEAL_KEY`) and writes a `*.credential`
+  file — which nothing in the mesh reads, and no consumer ever unseals.
+
+What the mesh already does, and has wired end to end:
+
+- The control plane mints one password per (consumer, provider) pair (`Inventory.SecretFor` →
+  `secrets.Make`) and seals it to **both** node keys asymmetrically — a copy the consumer's
+  host can open and a copy the provider's host can open. No shared symmetric key exists
+  anywhere, on purpose: a key both ends hold is a key the mesh would have to distribute, which
+  is the same problem one level down, and the control plane's own code refuses it.
+- The provider is handed, at the path its `receives` names, one contribution per consumer:
+  the **login to create** (`As`, derived by the mesh so the two ends agree by construction),
+  the consumer's address and requested values, and a **`Secret` file** holding that consumer's
+  password sealed to the provider and unsealed onto the machine by its host.
+- The consumer is handed the *same* password, as plaintext its own host wrote by unsealing its
+  copy and substituting it into a config file. The consumer never unseals anything itself and
+  holds no key.
+
+So the password a provider's provisioner invents is not even the password the consumer was
+given: a consumer authenticating with the mesh's password against a resource the provisioner
+created with its own would simply fail. The symmetric seal is not an incomplete feature to
+finish delivering a key for. It is a second, contradictory credential model bolted beside the
+real one, and it cannot be made to work without building the very thing the mesh was designed
+not to have.
+
+This is a decision and not a patch because the harness is the **provider contract**. Every
+provider — the four that exist and the many a real mesh grows — is built on
+`runProvisioner(resource, adapter)`. Whatever it says a provider is, they all inherit; and
+changing it later is one migration per provider. It is cheaper and more honest to settle what
+a provider is now.
+
+## Decision
+
+**A provider is handed the credential; it does not make one, does not seal one, and does not
+hand one back.** The provisioner's only job is to make the mesh's grants true in its own
+software.
+
+Concretely, for the sdk harness and the adapter contract:
+
+- The harness reconciles the **contributions the mesh delivers** to the provider's `receives`
+  path — the list of consumers, each with its login name (`As`), address, requested values,
+  and the path to its unsealed password (`Secret`). It does not read `*.grant.json` and it
+  does not write `*.credential`.
+- For each consumer present, the harness reads the password from that consumer's `Secret` file
+  and calls the adapter to bring the resource into being under the given login. For each
+  consumer no longer present — the mesh drops it from the contributions file when its consumer
+  goes away — the harness calls the adapter to withdraw it.
+- The adapter shrinks to the per-software half and nothing else. It is given the login, the
+  password, and the values, and it makes the resource exist or removes it. It generates no
+  password, derives no name, seals nothing, and returns no credential:
+  roughly `create({ as, password, values })` and `remove({ as })`, both returning nothing.
+- `$MESH_SEAL_KEY`, the symmetric `seal()`/`writeSealedCredential` path, and the `*.grant.json`
+  / `*.credential` files are removed from the provisioning path entirely. The credential
+  reaches the consumer through the mesh's own asymmetric channel, which already crosses node
+  boundaries and holds no shared secret.
+
+Identity stays the mesh's to say. The login the provider creates is the name the mesh derived
+and gave the consumer to present; the provider never invents a name, because a name the
+consumer cannot learn is a name it cannot authenticate with.
+
+## Consequences
+
+- A provider module becomes smaller and unable to be wrong in this way: with no password to
+  generate and no key to seal to, the class of bug where the two ends hold different secrets
+  cannot be written. A provider added after this inherits the corrected contract and has no
+  seal to reintroduce.
+- The four current providers (redis, postgres, minio, umami) each lose their `generatePassword`
+  + seal code and gain a `create` that takes the password it is given. Their teardown becomes
+  "withdraw the login named `As`".
+- The symmetric `seal()`/`unseal()` primitive loses its only caller and leaves — checked, not
+  assumed: nothing else in the sdk or the catalogue called it, so it is removed with the
+  provisioner it belonged to.
+- **How this is verified:** redis is assigned as a provider in the lab, the contributions and
+  the unsealed password the mesh would deliver are put in its `receives` path, and a client
+  authenticates as that consumer with the mesh's password and gets PONG — where a provider that
+  invented its own password answers WRONGPASS — with `$MESH_SEAL_KEY` set nowhere and no
+  `.credential` file written. Proven: `provider-uses-mesh-credential` is green.
+
+**What this does not cover — credential provisions, not data provisions.** This decision is about a
+provision whose credential is a *secret the mesh mints* — a login and password (redis, postgres,
+minio). A provider that instead *generates* the thing the consumer needs, and that thing is not a
+secret — umami's `analytics`, where the consumer wants back a `siteId` umami assigned — does not fit,
+because a contract that returns nothing has no way to hand that data back. The seal-key fault was
+never umami's (it sealed no password; it returned a public id), so removing the seal does not break
+it further, and it still reconciles its sites off the mesh's contributions. But delivering
+provider-generated data back to a consumer is a *return path* the mesh does not have and this
+decision does not build — a separate shape, left to a separate decision.
+- Teardown beyond "remove the login" — data an object store leaves behind when a consumer
+  leaves — is named by each provider's adapter, not by the harness, and is out of scope here
+  except to say the contract must leave room for it.
+
+## References
+
+- [04-ISSUES/032](../04-ISSUES/032-provider-runtime-has-no-seal-key/00-report.md) — the
+  observation and the cross-repo trace this decision rests on.
+- ADR 0047 (the module runtime) — what first ran a provider's provisioner as a delivered
+  process and exposed this; link to be filled when 0047 lands on the trunk.
+- Control-plane mechanisms this relies on already existing: `mesh-control` —
+  `internal/inventory/secrets.go` (`SecretFor`, `SecretsFrom`), `internal/secrets/seal.go`
+  (`Make`, the two-blob asymmetric sealing), `cmd/mesh-control/plan.go` (`grantsFor`, the
+  `Grant.Sealed = ForProvider` delivery), `internal/catalogue/declaration.go` (the `receives`
+  contribution: `As`, `At`, `Values`, `Secret`).
+- The path being removed: `mesh-sdk` — `src/provisioner/index.ts` (`runProvisioner`, `sealKey`,
+  `writeSealedCredential`) and the symmetric `src/primitives/index.ts` `seal()`/`unseal()`.

02-DECISIONS/0049-a-consumers-identity-fits-the-tightest-backend.md

@@ -0,0 +1,130 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-05
+deciders: jochen
+reconstructed: false
+---
+
+# 49. A consumer's identity is bounded by the tightest backend that must accept it
+
+## Context
+
+The mesh says who a consumer is, once, and hands the same name to the provider (to create) and the
+consumer (to present), so the two ends agree by construction rather than by two conventions (the
+principle behind `ConsumerIdentity`, 04-ISSUES/023). The name is `mesh_<node>_<module>`, cleaned to
+lower-case letters, digits and underscore.
+
+Proving the provider contract per backend (ADR 0048) turned up 04-ISSUES/034: redis and postgres
+create that name verbatim, but **minio refuses it** — an S3 access key is capped at 20 characters,
+and `mesh_anchor_bucketuser` is 22. The provisioner then retries for ever, per consumer, and the
+consumer holding that same too-long name could never present it either.
+
+Two things about the existing derivation decide most of this:
+
+- **The charset is already right.** `[^a-z0-9_]` is deliberately conservative, and its own comment
+  says it reaches "a PostgreSQL role, a MinIO access key, an LDAP uid and a Keycloak client without
+  quoting." That much is true.
+- **The length is wrong.** `CheckIdentity` refuses names over `identityLimit = 63`, commented as
+  "the shortest identifier limit among the systems these names reach: PostgreSQL's". It is not the
+  shortest — S3's 20 is shorter — so the guard that was meant to catch exactly this lets it through,
+  and the failure lands at provision time as a silent retry instead of at assignment as a refusal.
+
+So this is a small wrong constant with a real cost attached: whatever bound we set, `mesh_` (5) plus
+a node name plus `_` plus a module name has to fit inside it.
+
+## The options
+
+**A — Bound the identity by the true minimum, and refuse early.** Lower `identityLimit` to the real
+shortest (20, S3's), so `CheckIdentity` refuses an over-long name *at assignment* with a clear
+message, the way it already refuses over-63 names. The derivation does not change; long names are
+simply rejected before anything is provisioned.
+- *For:* smallest change; keeps "the mesh says the identity once, verbatim" intact; the failure
+  moves from a per-consumer provision-time retry to an up-front, legible refusal — which is what
+  `CheckIdentity` exists to do.
+- *Against:* a hard budget. `mesh_` + node + `_` + module ≤ 20 means node + module ≤ 14 characters.
+  `anchor` + `bucketuser` (16) is already over. It pushes the constraint onto how machines and
+  modules are named, which is a real limitation on legible names.
+
+**B — Keep the readable name when it fits, compact it when it does not.** Below the bound, the name
+is `mesh_<node>_<module>` as today; over it, the mesh substitutes a deterministic short form (e.g.
+`mesh_` + a truncated hash of node+module) — still one derivation, so both ends still agree.
+- *For:* no naming constraint; short backends always satisfied; the common case stays legible.
+- *Against:* some identities become opaque, and a provisioner tracing "whose login is this" loses
+  the answer for exactly the consumers that overflowed. The mesh now owns a fallback format and its
+  collision properties (a truncated hash is not free of collisions at 15 characters).
+
+**C — Let each interface declare its identifier bounds, and derive within the tightest a consumer
+reaches.** `s3-bucket` states `identifier: { max: 20 }`; `postgres-database` states 63; the mesh
+derives a name that fits the **minimum** bound across the providers a given consumer is granted.
+- *For:* the most precise — each provision gets exactly the room it has, and a database consumer
+  keeps long legible names while an S3 consumer gets a short one; the constraint lives where the
+  fact does (on the interface).
+- *Against:* the most work, and a consumer of two interfaces with different bounds must satisfy the
+  smaller — so its name shortens for both, reintroducing B's opacity in a narrower case. It also
+  means one consumer can hold **different** identities per provision, which the "said once" model
+  currently forbids.
+
+**D — Let the provider generate a backend-valid identity and hand it back (rejected).** minio mints
+its own access key and returns it to the consumer. This is the data-provision return path this era
+keeps meeting — but it directly contradicts 023 and ADR 0048: the identity would no longer be the
+mesh's single derivation the two ends share, it would be a value one side invents and the other must
+be told. Listed for completeness; not recommended.
+
+**E — A module (and a node) may declare a short slug; the identity is built from it.** The identity
+becomes `mesh_<node-slug|node-name>_<module-slug|module-name>`: where a slug is declared it is used,
+otherwise the cleaned name. A slug is a deliberately short, operator-chosen identifier — `kc` for
+keycloak, `wkstn` for a workstation. It is optional: short names (`anchor`, `redis`) need none.
+- *For:* this is the escape hatch B wanted to be, without the opacity. The name stays legible — a
+  provisioner can read `mesh_wkstn_kc` and know who is asking — because a person chose it, not a
+  hash function. And it makes an early refusal *palatable*: if even the slug-built identity overflows,
+  the refusal points at the slug, a field made for exactly this, rather than at the machine's name.
+  Both ends still derive it from one declared thing, so they agree by construction.
+- *Against:* a new optional manifest field, and someone must pick the slug — but only for names that
+  would otherwise overflow, and picking a short legible identifier is a better job than being handed
+  a hash.
+
+## What implementing A revealed
+
+A was tried first. At `identityLimit = 20`, the readable budget is `mesh_` (5) + node + `_` + module
+≤ 20, i.e. **node + module ≤ 14 characters** — far tighter than it looked. The catalogue's own
+existing tests use `workstation`+`keycloak` (25), which compacts to `mesh_dbbc02f8dde34d3`; common
+mesh names (`home-server`, `the-build-node`, `workstation`) blow the budget with any module. So B's
+compact fallback would fire for the *common* case, not the rare overflow — which inverts A+B: most
+identities would be opaque hashes. A alone (hard refusal at 20) would refuse most realistic names.
+This is what moved the recommendation to E: the problem is not the limit, it is that the *readable
+name* is the wrong source when it is long, and a slug is a better source than either a hash or a ban.
+
+## Recommendation
+
+**E, over a per-consumer bound (start with the global minimum, 20).** Build the identity from an
+optional slug, keep it when it fits, and refuse at assignment with "declare or shorten `<module>`'s
+slug" when it does not — no hash, no lost legibility, and the fix is a first-class field. Set the
+bound to the true minimum (20) now; it needs no per-interface machinery to unblock S3, and a module
+that consumes S3 simply declares a short slug. Graduate to **C** (per-interface bounds) later if it
+turns out that non-S3 consumers are paying for S3's limit often enough to mind — E and C compose:
+slugs are the mechanism, per-interface bounds refine where the ceiling sits. **B is dropped**: a
+declared slug is a strictly better escape hatch than an opaque hash. **D stays rejected.**
+
+## Consequences (of E)
+
+- A module manifest gains an optional `slug`; a node may carry one too. `ConsumerIdentity` prefers
+  the slug over the cleaned name for each half. `identityLimit` becomes 20 (the true minimum), and
+  `CheckIdentity` refuses at `module add` / assignment — now with a message naming the slug to set.
+- The common case stays legible; only names that overflow the budget need a slug, and what they get
+  is a name a person chose, not a hash.
+- Existing modules/nodes whose names overflow declare a slug once — a migration cost paid as a clear
+  refusal with an obvious remedy, not a silent hash or a silent truncation.
+- minio (04-ISSUES/034) is unblocked: an S3 consumer declares a short slug and its access key fits.
+- **How it is checked:** the minio grant e2e — a consumer whose (slugged) identity fits reaches its
+  bucket with the credential the mesh delivered — plus unit tests that a slug is preferred, that an
+  un-sluggable over-long identity is refused (naming the slug), and that two consumers never collide.
+
+## References
+
+- [04-ISSUES/034](../04-ISSUES/034-mesh-login-exceeds-s3-access-key-limit/00-report.md) — the
+  observation.
+- ADR 0048 — a provider creates the credential the mesh minted; the identity it creates it under is
+  the one this decision bounds.
+- `mesh-control` `internal/catalogue/identity.go` — `ConsumerIdentity`, `identityUnusable`,
+  `identityLimit`, `CheckIdentity` — where the constant and the check live.

02-DECISIONS/0050-model-access-is-vendor-agnostic.md

@@ -0,0 +1,206 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-05
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0024-model-access-is-a-provision.md
+---
+
+# 50. Model access is vendor-agnostic, and a vendor is an adapter
+
+## Context
+
+[ADR 0024](0024-model-access-is-a-provision.md) settled that model access is a provision and that
+a licence is a named thing an operator uses. What shipped, and runs, is a single vendor: the mesh's
+"claude" feature. A read-only trace of that feature (2026-09-05, in the code workspace) was made to
+answer whether the model-access provision is Anthropic-shaped or genuinely general. The finding is
+that **the vendor-agnostic layer already largely exists**, and the Anthropic specifics are a thin
+band around it that a per-vendor adapter can hold.
+
+**What is already general, with evidence.** `mesh-control internal/licences` models
+`licence(name, vendor, serves)` and `licence_holder(licence, node, module, sealed)`, and each
+holder's credential is sealed per-holder through `internal/secrets`. `serves` carries the
+non-secret facts (a base URL, a model) and is not vendor-specific. The `accept` verb
+([ADR 0024](0024-model-access-is-a-provision.md), and
+[`14-model-access`](../03-DESIGN/01-to-be/14-model-access.md)) already takes an operator-supplied
+value, seals it to each holder and discards the plaintext. A model the mesh runs itself answers
+`model-access` at node scope with no licence at all. None of that mentions Anthropic.
+
+**What is Anthropic-specific.** The credential is not a static key: it is a subscription OAuth grant
+— an hourly access token plus a refresh token. That shape drags four things behind it that a static
+key does not need: **central rotation** (one manager node refreshes under a lease and publishes the
+new token), **delivery that strips the refresh token** so a consuming node holds only an access token,
+an **identity guard** that reads the credential to catch a mis-binding, and a **usage** reading with
+Anthropic's own `utilization%` semantics. Most vendors are a single static key, which the sealed-key
+model already handles and which needs none of these four.
+
+**The tension at the centre of this.** A refreshable credential cannot be both *sealed so the mesh
+cannot read it* and *rotated centrally*. Central rotation means some node in the mesh holds the
+refresh token in readable form, because that is what refreshing requires. Per-holder sealing means no
+node but the holder can read the credential. For a static key the two never meet — there is nothing to
+rotate. For a refreshable grant they collide directly, and this record exists to say which gives way,
+and by how much.
+
+## Considered Options
+
+1. **Keep Anthropic special-cased in the core.** Leave the three binding columns and the `claude_*`
+   schema, and add other vendors beside them the same way. **Rejected.** It is exactly what
+   [ADR 0024](0024-model-access-is-a-provision.md) ruled against: a module that names a vendor cannot
+   be moved onto another model without editing it, and moving it is the point. It also grows the core
+   by one band per vendor, when the bands are the same shape.
+
+2. **One provision, and refuse to hold any refresh token — re-seal only.** Make every credential
+   purely sealed per-holder, including refreshable ones; let each holder refresh its own grant.
+   **Rejected.** It throws away the hard half [ADR 0024](0024-model-access-is-a-provision.md) says
+   already works — the lease, the single-refresher, the switch-on-exhaustion — and replaces it with N
+   nodes each holding a refresh token, which is the very thing today's delivery strips on the stated
+   ground that *a node never holds a refresh token*. A refresh token is the long-lived secret; spraying
+   it across every holder is strictly worse than keeping one copy on one node.
+
+3. **One provision, and abandon central rotation entirely** for refreshable vendors — treat the grant
+   as opaque and let it expire. **Rejected.** For a subscription-seat vendor an expired access token is
+   a dead licence; without rotation the feature that works today stops working. This is option 2's cost
+   without option 2's autonomy.
+
+4. **One vendor-blind provision, plus a per-vendor adapter, with a bounded carve-out for the
+   refreshable case.** **Adopted**, below.
+
+## Decision
+
+**`model-access` is one consumer-facing, vendor-blind provision.** A consumer declares
+`requires: model-access`, and is coupled to *reaching a model* — a base URL, a model name, a key —
+and not to which vendor answers. That is the coupling the name is drawn at
+([ADR 0040](0040-what-a-module-is.md)'s rule: name the interface at the widest boundary across which
+the consumer does not care which implementation serves it). Where a consumer were genuinely coupled to
+a specific wire API it could not swap across, the same rule would split the name — but the consumers
+that exist reach their model through a CLI or SDK that hides the vendor, so `model-access` is the true
+coupling and stays one name. This extends [ADR 0024](0024-model-access-is-a-provision.md) and
+[ADR 0027](0027-a-provision-names-what-the-consumer-is-coupled-to.md) without changing them.
+
+**A vendor is an adapter, keyed by the licence's `vendor` field.** The lifecycle a licence needs is
+vendor-specific and lives in a per-vendor adapter selected by `licence.vendor`, exactly as
+`public-dns` is one neutral interface answered by registrar-scoped providers —
+`cloudflare-dns`, `route53-dns` ([ADR 0044](0044-a-public-name-is-provisioned-like-any-capability.md)).
+A consumer names `model-access` and never a vendor, the same way a module names `public-dns` and never
+a registrar.
+
+**The field is named `vendor`, not `provider`.** The inventory already uses "provider" for the
+provider-pin — *which node answers a brokered provision*. Reusing it for *which company sells this
+licence* would collide two unrelated facts on one word. `vendor` is the licence's, and is separate.
+
+### The adapter's capabilities, all but one optional
+
+An adapter declares:
+
+- **`shape`** — `static-key` or `refreshable-grant`. This is the switch the carve-out below turns on.
+- **`accept(value) → sealed`** — take an operator-supplied credential and seal it to the holders, the
+  `accept` verb [ADR 0024](0024-model-access-is-a-provision.md) already defines.
+- **`refresh(licence)`** — refreshable-grant only: the lease / rotate / publish machinery.
+- **`identity(credential) → account-id`** — the mis-binding guard, for a vendor whose credential
+  carries an identity worth checking.
+- **`usage(licence) → normalised rows`** — the vendor's usage reading, mapped to the common shape below.
+- **`deliver`** — the credential *value* only; the destination path is the consumer's, not the
+  adapter's.
+
+**A static-key vendor implements almost nothing** — `shape: static-key`, `accept` is the generic
+seal, `deliver` is the value, and `refresh`, `identity` and `usage` are absent or trivial. The
+abstraction earns its keep by making the common vendor small, not the rare one clever.
+
+### The carve-out — the one place the guarantee is relaxed, said plainly
+
+The mesh's standing principle is that it cannot read what it stores: `accept` seals to the holders and
+discards the plaintext ([ADR 0024](0024-model-access-is-a-provision.md)), and a provider seals nothing
+because the credential travels the mesh's own asymmetric channel
+([ADR 0048](0048-a-provider-creates-the-credential-the-mesh-minted.md)). A `refreshable-grant`
+credential cannot honour that principle and be centrally rotated at the same time, and central rotation
+is the working half [ADR 0024](0024-model-access-is-a-provision.md) is explicit about keeping.
+
+**So, for `refreshable-grant` vendors only:**
+
+- the **manager node holds the refresh token encrypted at rest** — readable by that node, because
+  rotation requires it. This is the bounded exception.
+- **access tokens are still sealed per-holder**, as every credential is; a holder reads its own and no
+  other node reads it.
+- the **refresh token is stripped on delivery** — it never reaches a consuming node. *A node never
+  holds a refresh token* stays true for every node but the one manager.
+
+**Static-key vendors keep the full guarantee.** There is no token to rotate, so there is nothing to
+hold readably, so `accept` discards the plaintext and the carve-out never fires. The majority of
+vendors are static-key, and the majority therefore lose nothing.
+
+The exception is stated rather than hidden because a relaxed guarantee that is not written down is
+indistinguishable from a broken one. It is bounded on three axes at once: **refreshable-grant vendors
+only, the refresh token only, the manager node only.**
+
+### The settled details this record also fixes
+
+- **Usage is normalised to `(licence, consumer, period, metric, value)` plus the raw response as
+  jsonb.** The metric is vendor-defined — Anthropic's `utilization%` is one metric, a token count is
+  another — and no common unit is forced across vendors. The raw response is kept so a reading can be
+  re-derived if the normalisation is later found wrong.
+- **Binding is explicit per consumer, and an unchosen consumer is refused — no implicit fallback.**
+  This is the direction the resolver already takes, and it is the safe one: a mesh with several ways
+  to reach a model refuses a consumer that has not said which, naming the candidates and the command,
+  rather than silently choosing one ([ADR 0024](0024-model-access-is-a-provision.md), and
+  [`14-model-access`](../03-DESIGN/01-to-be/14-model-access.md)).
+- **Subscription-seat authentication lives entirely inside the adapter**, never in the generic core.
+  So does an interactive `/login` — an adapter-specific "adopt" origin for a credential a person must
+  produce in a browser; the generic `licence key <name>` covers the static-key case.
+- **Anthropic is the first `refreshable-grant` adapter**, carrying the OAuth refresh, the usage
+  reading, the identity guard and access-token-only delivery. **`anthropic-api-key` is a
+  `static-key` adapter for the same vendor's plain API keys**, and is the early second case that
+  proves the abstraction is not a single vendor wearing a coat: it exercises the whole path with the
+  carve-out switched off.
+
+## Consequences
+
+- **The carve-out is the mesh's one deliberate relaxation of "it cannot read what it stores."** It is
+  bounded to refreshable-grant vendors, to the refresh token, and to the manager node; the static-key
+  majority keep the full guarantee unchanged. This is the open risk the analysis carried here, and it
+  is recorded as an exception rather than pretended away.
+- **Anthropic collapses from special case to adapter.** The three binding columns
+  (`nodes.node_license`, `nodes.hal_claude_account`, `agents.claude_account`) become three ordinary
+  consumers of `model-access`; the `claude_*` schema becomes the generic licence tables plus one
+  adapter. What was hardcoded becomes data keyed by `vendor`.
+- **Adding a vendor is adding an adapter, and a static-key vendor is nearly free.** The modules that
+  want a model do not change when a vendor is added — they named `model-access`, not a vendor.
+- **The refresh-token concentration is now a stated property to defend, not an accident.** The manager
+  node is a place a long-lived secret lives readably, and losing it or compromising it is a bounded,
+  named blast radius rather than a surprise.
+
+### How each claim here is checked
+
+- **Vendor-blind provision, static-key path.** A lab scenario binds an `anthropic-api-key` licence to
+  a consumer; the consumer resolves, receives a key sealed to its node, and reaches a model — and the
+  key is **nowhere in the control plane's database** nor in anything that crossed the broker. This is
+  the `licence_holder` sealed-per-node check that [`14-model-access`](../03-DESIGN/01-to-be/14-model-access.md)
+  already runs, now asserted for a second vendor.
+- **The carve-out is exactly as narrow as stated.** For a `refreshable-grant` licence, a test asserts
+  the refresh token exists (encrypted) **only on the manager node**, is **absent from every holder's
+  delivery**, and that the delivered credential is access-token-only — and that for a `static-key`
+  licence no refresh token is stored anywhere.
+- **Adapter selection is keyed by `vendor`.** A scenario with two vendors on two licences verifies each
+  licence's lifecycle runs its own adapter, and that a consumer naming `model-access` never names a
+  vendor to get one.
+- **Refuse-if-unchosen.** Already checked in [`14-model-access`](../03-DESIGN/01-to-be/14-model-access.md):
+  a consumer with more than one candidate is refused with the candidates and the command named.
+
+## References
+
+- [ADR 0024](0024-model-access-is-a-provision.md) — model access is a provision, a licence is a named
+  thing, and `accept`; this record generalises its single vendor and keeps its working central
+  rotation.
+- [ADR 0027](0027-a-provision-names-what-the-consumer-is-coupled-to.md) — a provision names the
+  coupling; `model-access` is drawn at the consumer's.
+- [ADR 0040](0040-what-a-module-is.md) — the naming rule and the neutral-interface / scoped-provider
+  shape a vendor adapter follows.
+- [ADR 0044](0044-a-public-name-is-provisioned-like-any-capability.md) — registrar-scoped `public-dns`
+  providers, the precedent a `vendor`-scoped adapter mirrors.
+- [ADR 0048](0048-a-provider-creates-the-credential-the-mesh-minted.md) — the mesh seals credentials
+  and holds no readable copy; the carve-out here is the bounded, named exception to that for a
+  refreshable grant.
+- [`03-DESIGN/01-to-be/14-model-access.md`](../03-DESIGN/01-to-be/14-model-access.md) — the design this
+  record extends, amended to describe the adapter generalisation.
+- The read-only vendor-agnostic analysis, 2026-09-05 (code workspace) — the inventory and the decisions
+  taken on the open questions this record encodes.

02-DECISIONS/0051-shared-data-is-the-operators.md

@@ -0,0 +1,166 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-05
+deciders: jochen
+reconstructed: false
+extends: 0030-data-outlives-the-mesh-that-declared-it.md
+---
+
+# 51. Shared data is the operator's, and a module is granted access to it
+
+## Context
+
+**Eight modules declared one filesystem as eight private ones.** The media stack — a library
+server, the acquisition managers for films, series, music and books, a subtitle fetcher and two
+download clients — shares directories on one machine: the download clients write into
+`/services/media/downloads` and the managers read it; the managers write into the libraries and
+the library server reads them. That sharing is the entire point of the stack. Yet each module
+declared every shared directory it touched as its own `directory` resource, with an owner and a
+mode. `/services/media/downloads` was written seven times, as seven private directories that
+happen to be the same path.
+
+**The resolver refuses exactly that, and is right to.** Two modules declaring one path on one node
+are refused by name, with no exemption for identical content and no merge — because two owners of
+one path is the class of fault this repository keeps recording ([04-ISSUES/036](../04-ISSUES/036-six-modules-own-what-they-must-share/00-report.md)).
+So the stack as written refuses its own only sensible assignment: all of it on one machine,
+sharing one filesystem. It passed today only because no test co-resolves any two of the eight. The
+first machine assigned two of them together is where the refusal would have surfaced.
+
+**The vocabulary had one word for two intentions, and this was already seen.**
+[04-ISSUES/026](../04-ISSUES/026-the-data-directories-are-not-declared/00-report.md) found the
+same gap from the other side and named it precisely: two kinds of mount are spelled identically —
+*the directory my data lives in*, which the mesh creates and owns, and *a facility I was granted*,
+which already exists and the mesh only reaches. That issue deferred inventing a field to tell them
+apart, because doing so is a design decision and it declined to make one to get a check green. This
+is that decision.
+
+**A `directory` resource is owned, on every axis.** The host creates it, sets its owner and mode,
+and removes it when it is empty and no longer declared — it *removes what it made and leaves what
+it merely configured* ([ADR 0005](0005-the-node-host.md), [ADR 0030](0030-data-outlives-the-mesh-that-declared-it.md)).
+A shared media library is none of that. It existed before the mesh, several modules read and write
+it at once, and losing it is the one failure that does not recover. It is not any module's
+resource; it is the operator's, and a module only needs to be let at it.
+
+## Considered Options
+
+1. **Make the stack one module with several containers**, the way the mail system already is. The
+   issue raises it directly: is a set of modules that must share a filesystem really one module?
+   **Rejected.** The eight are independently assignable and independently useful — a person may run
+   the download client without the library server, or the film manager without the music one — and
+   folding them into a single module to express a shared directory would make *what a module is*
+   turn on an incidental filesystem contract. It also does not generalise: the next pipeline of
+   modules handing files to each other on one machine (an ingest folder, a spool, a drop directory)
+   would face the same wall and the same wrong remedy.
+
+2. **One module owns the directories and the rest `require` them.** **Rejected**, and this is the
+   heart of the decision. Nobody owns shared operator data. The library predates the mesh and
+   outlives any one module, so making the library server or a manager its owner means unassigning
+   that module orphans everyone else's access — and the owner would set the owner and mode of a
+   tree it did not create. Ownership is the wrong relationship to model, because the true owner is
+   not a module at all.
+
+3. **A flag on a directory resource** — `external: true`, or an owner of `operator`. **Rejected.**
+   It overloads one shape with a boolean that inverts every one of its semantics: created becomes
+   *must already exist*, owned becomes *touch nothing*, removed-when-empty becomes *never removed*.
+   That is the *two-kinds-spelled-identically* trap of 04-ISSUES/026 reintroduced with a single
+   quiet field — a reviewer reading `type: directory` would have to check one boolean elsewhere to
+   know whether the mesh owns the thing at all.
+
+4. **A distinct `accesses` declaration, separate from resources.** **Adopted.**
+
+## Decision
+
+**Shared, pre-existing data is operator-owned and external. The mesh does not create it, does not
+set its owner or mode, does not reconcile it and does not remove it.** A media library, a download
+spool, an ingest directory is the operator's, and the mesh is a guest in it.
+
+**A module declares that it needs *access* to such a path, not that it owns a resource there.** The
+manifest field is `accesses`: a list of `{path, mode}`, where mode is `read` or `read-write` and
+absent narrows to `read` — the safe default, because the danger with an access is being given more
+than was meant, not less. A module's own configuration and state directories stay owned
+`directory` resources; only the shared, pre-existing paths become accesses.
+
+**The host mounts an accessed path and owns nothing about it.** It reaches the machine as a new
+declaration shape, `access`, distinct from `directory`. The host confirms the path is present and
+does nothing else — no create, no chown, no mode, no removal.
+
+**An accessed path absent at apply time is refused, clearly, not created.** The mesh does not own
+it, so conjuring it would be a lie the host then acts on — and specifically the lie 04-ISSUES/026
+records, where a bind mount whose source does not exist is made by the container runtime as root
+with the wrong ownership. The host says the operator must provide the path instead.
+
+**Several modules accessing one path is normal, and never refused.** The duplicate-path refusal is
+about *ownership*, not *use*: it applies to resources a module owns and to those alone. An access
+is not a resource and never enters the check, so the eight-module stack co-resolves. What stays
+refused is genuine rivalry — two modules owning one path — and the new contradiction it exposes: a
+path one module owns while another merely accesses it, because that asserts both that the mesh owns
+the directory and that the operator does.
+
+This is a decision and not a patch because it settles *what a module may say about a path it did
+not make*, which every co-located file-handoff in the catalogue now and later depends on — and
+because it draws the ownership line [ADR 0030](0030-data-outlives-the-mesh-that-declared-it.md)
+started: the host owns what it made and keeps what it merely configured, and this adds the third
+case it did not have a word for — what it neither made nor configured, and must not touch.
+
+### How each claim is checked
+
+- **The stack co-resolves.** A control-plane unit test assigns two modules that declare access to
+  one path on one node and asserts no refusal — the exact case the resolver refuses when the same
+  path is owned. The mirror test, two modules *owning* one path, still refuses, so the sharing
+  vocabulary does not weaken the rule it sits beside.
+- **Ownership and access cannot both be claimed of one path.** A unit test asserts the resolver
+  refuses a path one module owns and another accesses, naming both.
+- **Absent is refused, not created.** A host unit test applies an access to a path that does not
+  exist and asserts a clear refusal that names the operator, and that nothing was created.
+- **Present is confirmed and nothing moves.** A host unit test applies an access to an existing
+  directory and asserts the apply reports no change and disturbs nothing.
+- **Undeclaring never removes.** A host unit test drops a previously declared access and asserts
+  the operator's directory and its contents are left exactly as they were — the data-loss failure
+  [ADR 0030](0030-data-outlives-the-mesh-that-declared-it.md) exists to prevent, on a directory the
+  mesh never made.
+- **Every media manifest is corrected.** No `/services/media/*` path is an owned `directory`
+  resource in any of the eight; each is an `accesses` entry, and each module's own config and state
+  directories remain owned. Checked by the control-plane manifest parser, which now understands
+  `accesses` and refuses a malformed one.
+
+## Consequences
+
+**A shared filesystem between co-located modules now has a vocabulary**, and it is not the media
+stack's alone: any pipeline handing files to a neighbour on one machine — an ingest directory, a
+spool, a drop folder — says *I access this operator path* rather than *I own this directory*, and
+several of them may say it of one path.
+
+**Unassigning a module that reached shared data leaves the data.** Correct, and the same trade
+[ADR 0030](0030-data-outlives-the-mesh-that-declared-it.md) made for owned directories: removing
+data is a person's act, done knowingly, not a side effect of unassignment.
+
+**The operator must provision the shared paths before the stack is applied**, and a machine that
+lacks one is told plainly which. That is a real new obligation, and it is the right one: the mesh
+cannot own what predates it, so it cannot create it either, and saying so at apply time beats a
+directory conjured as root and a service that half-works.
+
+**The host vocabulary grew by one shape**, which is a cost — every added shape widens what a
+compromised control plane can express ([ADR 0005](0005-the-node-host.md)). It is a narrow one: an
+`access` is confirmed by a stat and grants the host no new action. It earns its place by letting
+the host refuse to create what it must not own, which no existing shape could say.
+
+**A path can be both owned and accessed only by refusal.** If a future manifest declares one path
+as an owned directory in one module and an access in another, the resolver refuses it rather than
+guessing which is meant — the two assertions about who owns the data cannot both hold.
+
+## References
+
+- [04-ISSUES/036](../04-ISSUES/036-six-modules-own-what-they-must-share/00-report.md) — six (in
+  fact eight) modules own what they must share; the problem this resolves
+- [04-ISSUES/026](../04-ISSUES/026-the-data-directories-are-not-declared/00-report.md) — the two
+  kinds of mount spelled identically, which deferred this field to a decision
+- [ADR 0030](0030-data-outlives-the-mesh-that-declared-it.md) — data outlives the mesh; the host
+  keeps what it did not make. This record adds the case it had no word for
+- [ADR 0005](0005-the-node-host.md) — the host removes what it made and leaves what it merely
+  configured; the vocabulary is finite and every shape is a security decision
+- [ADR 0040](0040-what-a-module-is.md) — what a module is; an access is a new thing a module may
+  say about the machine it lands on
+- mesh-control `feat/shared-data-access`, mesh-catalog `feat/media-access-not-ownership`,
+  mesh-host `feat/mount-operator-owned` — the mechanism, the corrected manifests, and the host
+  shape

02-DECISIONS/0052-a-step-that-runs-once-before-a-container.md

@@ -0,0 +1,198 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-05
+deciders: jochen
+reconstructed: false
+extends: 0005-the-node-host.md
+---
+
+# 52. An init step is a container run once to completion, gating what follows
+
+## Context
+
+**A module can declare things that exist; it cannot declare a step that runs.** The host owns a
+finite vocabulary of shapes — `file`, `directory`, `service`, `package`, `container`, `action` —
+and every one but `action` describes *state*: a thing that should be present, with content or a
+mode or an image, which the host reconciles toward ([ADR 0005](0005-the-node-host.md)). That is
+right for what it covers. But a real class of modules needs, once, to *run their own code at a
+point in their own lifecycle* — and the vocabulary has no word for it
+([04-ISSUES/037](../04-ISSUES/037-a-module-cannot-run-code-at-a-lifecycle-phase/00-report.md)).
+
+**mosquitto is the sharp case, and it fails silently without this.** Its Dynamic Security plugin
+loads at broker start and refuses to come up unless `dynamic-security.json` already holds an admin
+client. Seeding that file is a step that must happen *after* the data directory exists and *before*
+the broker container starts. The manifest can declare the directory, the config file and the broker
+container; it cannot declare "seed this, once, before that container starts." Written as it is
+today, the broker starts against an unseeded store and the plugin aborts — and the next reconcile
+does not fix it, because nothing in the declaration ever seeds the file.
+
+**It is not one module's defect.** The database providers need the same to run a first-boot
+migration, an extension enable, or a health gate before they are announced ready; today that works
+only where the *image* happens to seed itself from an environment variable, and anything the mesh
+must run once against the server has no home. This is the timing face of the same gap
+[04-ISSUES/035](../04-ISSUES/035-reconciling-a-seed-file-wipes-what-grew-in-it/00-report.md) records
+from the content side: the manifest needed *the file to exist before first start*, and had only
+*the file has this content, forever*.
+
+**The obvious answer is the one that already went wrong.** An earlier mesh had exactly this as a
+feature — event-driven hooks that ran custom code at phases of build, publish and deploy. It was
+powerful and it was *complex to set up and flaky*, and that fragility, not the need, is the content
+of the issue. Whatever this becomes must not rebuild that engine.
+
+**The ground has shifted since that engine, in a way that makes a much smaller answer possible.** A
+module with tools or events now runs a **process of its own** — a container carrying the module's
+compiled code, holding the single broker account the mesh scoped to it, isolated from every other
+module ([ADR 0047](0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md)). The
+code that must run at first boot is *already in that image, under that account*. So the mesh does
+not need a way to run a module's code — it has one. It needs a way to say **run this container to
+completion, and start the one that depends on it only after it has.**
+
+## Considered Options
+
+1. **A host `run` shape — a command the host executes on the machine.** The direct reading of
+   "run code at a lifecycle phase." **Rejected.** The host's vocabulary is finite and every added
+   shape is a security decision, because it widens what a *compromised control plane* can express
+   ([ADR 0005](0005-the-node-host.md)). A general "run this command as the host" is the largest
+   such widening there is: the blast radius is the whole machine, as root. The mesh already drew
+   this exact line for `action` — it runs a command, and so it is *permitted from the bundle and
+   refused from the link*, because the bundle arrives with the binary and the link is a separate
+   party with an unbounded reach. A new host-command shape usable by an ordinary module would be an
+   `action` from the link by another name, which is precisely what is refused.
+
+2. **Per-phase lifecycle hooks on a module** — `pre-start`, `post-start`, `pre-remove`, and their
+   build/publish cousins, each naming code the mesh runs at that phase. The general answer, and the
+   old feature. **Rejected for now.** It is the flaky engine the issue warns against, and most of
+   its phases have no present need. Deciding the full set of phases, where each one's code runs, and
+   how each is made idempotent is a large design taken to buy capability nothing yet asks for. The
+   three blocked modules all need one phase — *before a container starts* — and a mechanism narrow
+   enough to be obviously correct beats a general one that is not.
+
+3. **A distinct one-shot resource type** — a new shape, sibling to `container`, that names an image
+   and runs it once. **Rejected.** It grows the host vocabulary by a whole shape (a `Type`, a
+   struct, an applier, a place in every host's shape list) to express something a `container` almost
+   already is. A one-shot *is* a container — a pinned image, an account, volumes, an environment —
+   that happens to exit. Spending a new shape on the difference is the cost of option 1 in smaller
+   type, for a capability the existing shape can carry with one modifier.
+
+4. **A modifier on the existing `container` shape: this container runs once, to completion, and the
+   host gates the apply on it.** **Adopted.** It reuses the shape the host already has, adds no new
+   host action, and leans on two guarantees the host already gives — *apply in declared order,
+   never sorted*, and *a failed step fails the apply* — to turn "before that container starts" into
+   an emergent property of ordering rather than a dependency graph the host must resolve.
+
+## Decision
+
+**A run-once step is an ordinary `container`, marked to run to completion.** The manifest sets
+`run-once: true` on a container resource. Everything else about it is a container as before — a
+digest-pinned image ([ADR 0006](0006-the-substrate-and-the-control-plane.md)), volumes, an
+environment, and for a module's own code the same scoped account its runtime already holds
+([ADR 0047](0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md)). The mesh adds
+no way to run code; it marks a container the host must **run to completion and require to exit 0**,
+rather than start and leave running.
+
+**The gate is declaration order, not a named dependency.** The host applies a declaration in the
+order it is given, does not sort, and does not resolve dependencies — ordering is a decision, and it
+is the control plane's ([ADR 0005](0005-the-node-host.md),
+[04-ISSUES/013](../04-ISSUES/013-a-file-arrives-after-the-service-that-needs-it/00-report.md)). A
+run-once step is placed *before* the container that depends on it, and **a failed run-once step
+halts the apply**, exactly as a failed `action` does — so everything the declaration places after
+it, the broker included, is never reached until the step has completed. "Before the broker starts"
+is therefore expressed by list position plus completion, and the host cross-references nothing.
+
+**Completion is recorded, and a re-apply does not re-run it.** The host records what it applied only
+after the fact, as the digest of the declaration that produced it
+([ADR 0018](0018-a-picture-is-read-from-what-runs.md)) — a run-once step no differently. Because the
+step leaves nothing running to inspect, that persisted digest, not a live container, is the marker
+that it happened. On a later apply the host finds the digest already recorded for this exact
+declaration and does nothing; it re-runs only when the declaration's digest has changed, and a step
+that exited non-zero recorded nothing and so is retried next apply. This is the reconcilable,
+idempotent discipline the state shapes get for free, made explicit for a step.
+
+This is a decision and not a patch because it settles **what a module may say about running its own
+code**, which the whole catalogue of providers — a seed before start, a first-boot migration, a
+health gate — now and later depends on, and because it draws the line the issue asked for: the
+narrowest sound mechanism that unblocks the three modules without rebuilding the hook engine whose
+fragility is the warning.
+
+### The security bound, stated plainly
+
+**The host gains no new action and no new shape.** `run-once` is a boolean modifier on the
+`container` shape that already exists. A run-once container is strictly *less* powerful than an
+`action`: it cannot run an arbitrary host command, only a digest-pinned image under an account the
+mesh scoped — which is exactly the capability `container` already grants from the link. A control
+plane that is compromised can express nothing through `run-once` it could not already express by
+declaring an ordinary `container`. The dangerous expansion of option 1 — a command the host runs on
+the machine — is not made.
+
+### How each claim is checked
+
+- **A run-once step runs to completion and its exit 0 is required.** A host unit test applies a
+  run-once container whose image exits 0, asserts the host ran it to completion (not detached, not
+  left running) and reported it done; a sibling test applies one that exits non-zero and asserts
+  the apply fails, naming the step.
+- **A failed run-once step gates what follows.** A host unit test places a run-once container that
+  exits non-zero before another container and asserts the second is never started and the apply is
+  reported gated — the mirror of the existing test that a failed action stops what follows.
+- **It is not re-run once it has completed.** A host unit test applies a run-once step, then applies
+  the identical declaration again with the first run's record present, and asserts the second apply
+  runs nothing and reports the step unchanged.
+- **A changed declaration re-runs it.** A host unit test applies a run-once step, then applies one
+  whose image or environment differs, and asserts it runs again — the digest moved, so the marker no
+  longer matches.
+- **The vocabulary carries the field end to end.** A control-plane unit test resolves a module whose
+  manifest marks a container `run-once` and asserts the rendered host declaration carries the field,
+  in author order before the container it gates; the manifest parser refuses a `run-once` that is
+  not boolean.
+- **mosquitto seeds before the broker.** mosquitto's manifest declares a run-once init container,
+  before the `server` (broker) container, that writes the admin client into `dynamic-security.json`
+  and exits — checked by the control-plane resolver, which now understands the field, and by the
+  ordering of the rendered declaration. The end-to-end proof that it runs exactly once, at the right
+  phase, and converges on re-apply is owed to a lab scenario ([04-ISSUES/037] open question), which
+  this record does not close.
+
+## Consequences
+
+- **The three blocked modules gain a home for their step.** mosquitto seeds its dynsec admin before
+  the broker; a provider that must migrate or health-gate at first boot declares a run-once step in
+  its own runtime image, under its own account, before the container that depends on it.
+- **The general lifecycle hook is deferred, deliberately.** Only *before a container starts* is
+  bought here. `post-start`, `pre-remove` and the build/publish phases remain unbuilt, and the day
+  one is genuinely needed it is decided then, against a need, not speculatively — the same restraint
+  that kept this from being the old engine.
+- **The seed-then-mutate file is safe if the step is written to be.** A run-once seed writes
+  `dynamic-security.json` only when it is absent and never reconciles it, so what the running plugin
+  grows in that file afterward is never wiped
+  ([04-ISSUES/035](../04-ISSUES/035-reconciling-a-seed-file-wipes-what-grew-in-it/00-report.md)).
+  The host's marker guarantees the step is not re-run; the step's own code guarantees it does not
+  clobber on the pass it does run.
+- **The host vocabulary did not grow, and that is the point.** The cost of a run-once step is one
+  boolean and a completion path in the container applier, not a new shape and not a new action. The
+  mesh expresses ordering and completion; the module runs its own code, where it already runs it.
+- **A run-once step that never converges is a stuck apply, loudly.** A step that exits non-zero
+  every time halts the apply every time, and the container it gates never starts — which is the
+  correct failure, reported, rather than a broker that half-starts against an unseeded store and a
+  reconcile that reports success. It is failed forward, not failed silent.
+
+## References
+
+- [04-ISSUES/037](../04-ISSUES/037-a-module-cannot-run-code-at-a-lifecycle-phase/00-report.md) — a
+  module cannot run its own code at a lifecycle phase; the gap this resolves, and the warning about
+  the old hook engine
+- [04-ISSUES/035](../04-ISSUES/035-reconciling-a-seed-file-wipes-what-grew-in-it/00-report.md) — the
+  content face of the same gap: a file needed before first start, that the running program then
+  mutates
+- [04-ISSUES/013](../04-ISSUES/013-a-file-arrives-after-the-service-that-needs-it/00-report.md) — the
+  order of a declaration is the control plane's, and the host applies it as given; the gate rests on
+  this
+- [ADR 0005](0005-the-node-host.md) — the host's finite vocabulary, the ordered declaration it does
+  not sort, and `action` as the shape a command already is and why it is refused from the link
+- [ADR 0047](0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md) — a module runs
+  its code as its own process under its own account; a run-once step is that process, run to
+  completion
+- [ADR 0018](0018-a-picture-is-read-from-what-runs.md) — what was applied is recorded after it works;
+  the completion marker is that record
+- [ADR 0006](0006-the-substrate-and-the-control-plane.md) — a container is pinned by digest; a
+  run-once container no differently
+- mesh-control `feat/lifecycle-run-once`, mesh-host `feat/apply-run-once`, mesh-catalog
+  `feat/mosquitto-bootstrap` — the vocabulary, the apply support, and mosquitto's seeded broker

02-DECISIONS/0053-a-step-that-runs-on-a-schedule.md

@@ -0,0 +1,181 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-06
+deciders: jochen
+reconstructed: false
+extends: 0052-a-step-that-runs-once-before-a-container.md
+---
+
+# 53. A scheduled step is a container run on a recurring schedule
+
+## Context
+
+**[ADR 0052](0052-a-step-that-runs-once-before-a-container.md) gave the mesh a step that runs *once*;
+a real class of modules needs one that runs *again and again*.** kometa reconciles a media library
+against its lists on a timer; a ticketing integration polls its source for new work every few minutes;
+a backup, a cache warm, a metrics roll-up all recur. The host's vocabulary describes *state* — a file,
+a directory, a container that should be running — and 0052 added *a step that happens once and is
+done*. Neither says *this should happen every night at 3, forever*
+([04-ISSUES/037](../04-ISSUES/037-a-module-cannot-run-code-at-a-lifecycle-phase/00-report.md) named
+the lifecycle gap; 0052 closed the once-before-start face of it and left the recurring face open).
+
+**The modules that need it already run their own code.** As with run-once, the ground has shifted
+since the old mesh's flaky hooks: a module with tools or events runs a **process of its own** — a
+container carrying the module's compiled code under the single scoped account the mesh gave it
+([ADR 0047](0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md)). The work that
+must recur is already in that image, under that account. The mesh does not need a way to run a
+module's code on a timer — it has the code and the account. It needs a way to say **run this
+container again on this cadence**.
+
+**The shape of the answer is already decided, one modifier over.** 0052 rejected a host `run` command,
+per-phase hooks, and a distinct one-shot resource type, and adopted *a modifier on the `container`
+shape the host already has*, because a one-shot is a container that happens to exit. A scheduled step
+is the same container that happens to exit — run repeatedly. Reusing that shape keeps the host
+vocabulary flat and inherits 0052's security bound whole. The only thing 0052's `run-once` does not
+carry is *when to run it again*.
+
+**One difference from run-once changes a rule, and it is the reason this is its own record.** A
+run-once step **gates the apply**: it is placed before the container that depends on it, and a failure
+halts everything after it, because "seed the store before the broker starts" is a correctness
+precondition ([ADR 0052](0052-a-step-that-runs-once-before-a-container.md)). A scheduled step is the
+opposite: it runs *after* the machine is up and converged, on its own clock, and a single failed run
+is an ordinary operational event — the next run comes anyway. A scheduled step that halted the apply,
+or that a failed run marked the node not-current over, would make a routine poll into a reason the
+whole machine reads as broken. So the gating rule 0052 established is exactly the rule this record
+must **not** inherit.
+
+## Considered Options
+
+1. **A host `cron`/`timer` shape — the host installs a system timer that runs a command.** The direct
+   reading. **Rejected**, for the reason 0052 rejected a host `run` shape: it widens what a
+   *compromised control plane* can express toward "run this command on the machine, forever," which is
+   the largest widening there is, and it is `action`-from-the-link by another name
+   ([ADR 0005](0005-the-node-host.md)). A recurring command is worse than a one-off, because it
+   persists.
+
+2. **Per-phase lifecycle hooks** — `on-schedule` joining `pre-start`/`post-start` as named code the
+   mesh runs. **Rejected for now**, as in 0052: it is the flaky hook engine the issue warns against,
+   and the three modules that need this need one thing — *run this container on a cadence* — which a
+   narrow modifier expresses without deciding a whole hook vocabulary.
+
+3. **A distinct `scheduled` resource type**, sibling to `container`. **Rejected**, as 0052 rejected a
+   distinct one-shot type: it spends a whole new host shape (a `Type`, a struct, an applier, a place
+   in every host's shape list) on something a `container` already almost is — a scheduled task *is* a
+   container (pinned image, account, volumes, environment) that runs on a clock.
+
+4. **A modifier on the existing `container` shape: `schedule`, a cron expression the host runs the
+   container on.** **Adopted.** It reuses the shape the host has, adds no new host action, and sits
+   beside `run-once` as its recurring twin — the same container, exited, run again.
+
+## Decision
+
+**A scheduled step is an ordinary `container`, marked with a `schedule`.** The manifest sets
+`schedule: "<cron>"` on a container resource — a standard five-field cron expression. Everything else
+about it is a container as before: a digest-pinned image
+([ADR 0006](0006-the-substrate-and-the-control-plane.md)), volumes, an environment, and for a module's
+own code the same scoped account its runtime already holds
+([ADR 0047](0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md)). The mesh adds no
+way to run code; it marks a container the host must **run on that cadence, each time to completion**,
+rather than start once and leave running (a service) or run once and gate (a run-once step).
+
+**A run is fired by the clock, not by the apply, and does not gate it.** Applying the declaration
+installs the schedule; it does not run the step. The machine converges — reports applied and current —
+as soon as the schedule is installed, exactly as it does for a service that is running. Thereafter the
+host fires the container when the cron expression is due. This is the deliberate inversion of 0052:
+a scheduled step is downstream of convergence, not a precondition of it.
+
+**A failed run is recorded and the next run still comes; it never marks the node not-current.** A run
+that exits non-zero is logged against the module — visible, auditable — but it does not fail the apply,
+does not halt other resources, and does not flip the node's reported state. A poll that fails at 03:00
+and succeeds at 03:05 is the system working, not a machine that needs attention. A step that fails
+*every* time is a loud, repeating log entry, which is the correct signal for "this recurring job is
+broken" — distinct from "this machine did not converge."
+
+**Runs do not stack.** If a run is still going when the next is due, the host skips the due run rather
+than starting a second copy, and logs the skip. A slow nightly job that occasionally overruns must not
+spawn a growing pile of concurrent containers competing for the same account and volumes — the failure
+mode that made the old timers dangerous.
+
+**Each run is independent and idempotent by the module's own code.** The mesh guarantees only *the
+container is run on the cadence*; that a run does the right thing when the previous one half-finished
+is the module's contract, the same discipline a run-once seed owes
+([ADR 0052](0052-a-step-that-runs-once-before-a-container.md),
+[04-ISSUES/035](../04-ISSUES/035-reconciling-a-seed-file-wipes-what-grew-in-it/00-report.md)).
+
+This is a decision and not a patch because it settles **what a module may say about running its own
+code on a cadence**, which every recurring provider job — a sync, a poll, a roll-up — now and later
+depends on, and because it draws the line 0052 left: the recurring twin of run-once, with the gating
+rule deliberately reversed so a routine job's failure is never a machine's failure.
+
+### The security bound, stated plainly
+
+**The host gains no new action and no new shape.** `schedule` is a string modifier on the `container`
+shape that already exists. A scheduled container is strictly *less* powerful than an `action`: it runs
+a digest-pinned image under an account the mesh scoped, which is exactly what `container` already
+grants from the link, and it cannot run an arbitrary host command. A compromised control plane can
+express nothing through `schedule` it could not already express by declaring a `container` — the cron
+string only says *how often*, not *what*. The dangerous expansion of option 1 — a command the host
+runs on the machine on a timer — is not made.
+
+### How each claim is checked
+
+- **A scheduled step runs when the schedule is due.** A host unit test installs a container with a
+  schedule that is due immediately (or advances a injected clock to when it is due) and asserts the
+  host ran it to completion; a sibling test with a schedule not yet due asserts it has not run.
+- **Installing it does not run it, and the node is current without a run.** A host unit test applies a
+  scheduled container and asserts the apply reports current *before* any run has fired — the schedule
+  is state that is present, not a step that gated.
+- **A failed run does not fail the apply or the node.** A host unit test fires a scheduled container
+  that exits non-zero and asserts the failure is recorded against the module, the apply is not failed,
+  and the node stays current — the mirror of the run-once test where a non-zero exit *does* halt.
+- **Runs do not stack.** A host unit test fires a scheduled container whose run outlasts its next due
+  time and asserts the host skipped the due run and logged the skip, rather than starting a second
+  container.
+- **The vocabulary carries the field end to end.** A control-plane unit test resolves a module whose
+  manifest sets `schedule` on a container and asserts the rendered host declaration carries the field;
+  the manifest parser refuses a `schedule` that is not a valid cron expression, and refuses a container
+  that is both `run-once` and `schedule` (a step is one or the other, never both).
+- **A real module recurs in the lab.** A converted module declaring a scheduled step (kometa's library
+  sync, or a poller) is assigned in a lab scenario, and the scenario asserts the scheduled container
+  fires on its cadence and its account and volumes are the module's — the end-to-end proof this record
+  owes, as 0052 owed its run-once lab proof.
+
+## Consequences
+
+- **The recurring providers gain a home for their cadence.** kometa reconciles on its schedule; a
+  poller polls; a roll-up rolls up — each a scheduled step in its own runtime image, under its own
+  account, on the cron it declares.
+- **The general lifecycle hook is still deferred.** Only *run once before* (0052) and *run on a
+  cadence* (this) are bought. `post-start`, `pre-remove` and the build/publish phases remain unbuilt,
+  decided when a real need arrives, not speculatively — the restraint that kept both from being the old
+  engine.
+- **A recurring job's failure is loud but not fatal.** The node stays current while a scheduled step
+  fails and retries; a step that fails forever is a repeating log entry, not a machine marked broken.
+  This is the correct separation — a machine's convergence and a job's success are different questions —
+  and it is why this could not simply be `run-once` without the schedule.
+- **The host vocabulary did not grow, again, and that is the point.** The cost of a scheduled step is
+  one string field and a cron loop in the container applier, not a new shape and not a new action. The
+  mesh expresses cadence; the module runs its own code, where it already runs it.
+- **`run-once` and `schedule` are exclusive and complete for now.** A container runs once and gates, or
+  runs on a cadence and does not, or runs and stays up (a service). A manifest that asks for two of
+  these at once is refused, because the three are distinct answers to "how does this container run."
+
+## References
+
+- [ADR 0052](0052-a-step-that-runs-once-before-a-container.md) — the run-once step; this is its
+  recurring twin, reusing the `container`-modifier shape and inheriting its security bound, and
+  reversing its gating rule
+- [04-ISSUES/037](../04-ISSUES/037-a-module-cannot-run-code-at-a-lifecycle-phase/00-report.md) — a
+  module cannot run its own code at a lifecycle phase; 0052 closed the once-before-start face, this
+  closes the recurring face
+- [ADR 0047](0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md) — a module runs
+  its code as its own process under its own account; a scheduled step is that process, run on a cadence
+- [ADR 0005](0005-the-node-host.md) — the host's finite vocabulary, and why a host command (even on a
+  timer) is refused from the link
+- [ADR 0006](0006-the-substrate-and-the-control-plane.md) — a container is pinned by digest; a
+  scheduled container no differently
+- [ADR 0018](0018-a-picture-is-read-from-what-runs.md) — what was applied is recorded after it works;
+  the installed schedule is state, a fired run is an event
+- mesh-control `feat/schedule-container`, mesh-host `feat/apply-schedule`, and the converted module
+  that first declares a scheduled step — the vocabulary, the apply support, and the recurring proof

02-DECISIONS/0054-model-usage-is-recorded-at-two-grains.md

@@ -0,0 +1,178 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-06
+deciders: jochen
+reconstructed: false
+extends: 0050-model-access-is-vendor-agnostic.md
+---
+
+# 54. Model usage is a vendor-neutral record, produced by the adapter, at two grains
+
+## Context
+
+**[ADR 0050](0050-model-access-is-vendor-agnostic.md) fixed the *shape* of a usage reading and left
+its *home* open.** It decided that an adapter may expose `usage(licence) → normalised rows`, that a
+row is `(licence, consumer, period, metric, value)` plus the raw vendor response as jsonb, and that the
+metric is vendor-defined (Anthropic's `utilization%` is one metric, a token count another). It did not
+say where those rows are stored, how they get produced on a cadence, or whether the only grain is the
+whole licence — and the mesh's first real consumer needs all three answered.
+
+**The predecessor mesh recorded usage at two grains, and both are wanted.** It polled the vendor for a
+licence-level reading (an account's `utilization%`), and it also attributed **per-session** token and
+cost — which model, how many input and output tokens, what it cost — parsed from the agent's own
+transcript, including the case where a long session switched the account it billed against mid-way. The
+licence-level reading answers "how close is this subscription to its cap"; the session-level reading
+answers "what did this piece of work cost, and against which account." A mesh that kept only the first
+could not bill a project or notice a runaway session; keeping only the second could not see a cap
+approaching. Both are load-bearing and neither subsumes the other.
+
+**A session is already a consumer, so the second grain needs no second vocabulary.**
+[ADR 0026](0026-the-mesh-has-a-session-of-its-own.md) and design 15 establish that an agent session is
+a thing in its own right, and that **model access is bound to the session, not to the machine** — a
+session is a *consumer* of `model-access`, identified by `(node, module)` and its own id. That is
+exactly the `consumer` column ADR 0050 already put in the usage row. So the two grains are not two
+schemas; they are the same row at two consumer resolutions: the holding module for the licence grain,
+the session for the finer one. The design's own still-open worker-naming gap (design 14) is the same
+gap here and is left where it is — a session id distinguishes what `(node, module)` cannot.
+
+**The pieces this needs already exist.** A periodic reading is a **scheduled step**
+([ADR 0053](0053-a-step-that-runs-on-a-schedule.md)) — the adapter's `usage()` poll is a container the
+mesh runs on a cadence, which is precisely what 0053 was built for. A durable audit of what happened is
+**an event the audit trail records** ([ADR 0041](0041-events-are-a-relationship.md),
+[ADR 0042](0042-the-shape-of-an-event-on-the-wire.md)); the `audit-logger` module already consumes every
+event. And a queryable current picture is **a context store**, the same shape the licences themselves
+live in ([ADR 0008](0008-a-context-owns-its-store.md)). Nothing new in kind is required; what is missing is
+the decision to point them at usage.
+
+## Considered Options
+
+1. **Licence grain only — a single `utilization%` poll, nothing per session.** Rejected: it cannot
+   attribute cost to a piece of work or catch a session that is burning an account down, which is half
+   of why usage is recorded at all.
+
+2. **A bespoke `sessions` schema mirroring the old mesh's `*_sessions` / `*_session_account_usage`
+   tables.** Rejected: it reintroduces a second, vendor-shaped vocabulary for something the mesh
+   already names — a session is a consumer, and its usage is a usage row. A parallel schema would drift
+   from the `model-access` vocabulary and force every reader to learn two.
+
+3. **Store usage only as raw vendor blobs, normalise later.** Rejected as the *whole* answer (kept as a
+   fallback within the chosen one): a reader that must parse Anthropic's response shape to answer "what
+   did this cost" has the vendor coupling the whole feature exists to remove. The raw blob is kept
+   beside the normalised row (0050 already requires this), not instead of it.
+
+4. **One vendor-neutral usage record at two consumer grains, produced by the adapter, recorded as
+   both an event and a queryable row.** Adopted.
+
+## Decision
+
+**Model usage is one vendor-neutral record — `(licence, consumer, period, metric, value)` plus the raw
+response — recorded at two grains that differ only in the `consumer`.** At the **licence grain** the
+consumer is the holding module and the metric is the vendor's own account reading (Anthropic:
+`utilization%`). At the **session grain** the consumer is the agent session — `(node, module)` and its
+session id ([ADR 0026](0026-the-mesh-has-a-session-of-its-own.md)) — and the metrics are the ones a
+session bills: input tokens, output tokens, model, and cost. The row shape is 0050's, unchanged; the
+grain is which consumer the row is *for*.
+
+**The adapter is the only thing that knows the vendor, and it produces both grains.** Reading an
+account's cap is the adapter's `usage(licence)` verb ([ADR 0050](0050-model-access-is-vendor-agnostic.md));
+attributing a session's cost is the adapter reading that vendor's transcript or usage API and emitting
+rows keyed to the session. The mesh defines the row and the plumbing; the adapter fills it from whatever
+the vendor exposes, and a static-key vendor that exposes nothing simply produces no rows — usage is an
+optional reading, not a requirement of holding a licence.
+
+**A reading is taken on a schedule, not on a request.** The licence-grain poll is a **scheduled
+container** ([ADR 0053](0053-a-step-that-runs-on-a-schedule.md)) the adapter runs on a cadence; the
+session-grain rows are produced as sessions progress, from the transcript the session already writes.
+Neither blocks anything: a poll that fails is a logged, retried scheduled run (0053's rule), and a
+session whose cost cannot yet be attributed is a row not yet written, never a session refused.
+
+**Usage is recorded two ways, for two audiences.** Each reading is **emitted as an event**
+([ADR 0041](0041-events-are-a-relationship.md)) — an immutable "this was observed at this time" that the
+`audit-logger` already records, so the history of what an account did is in the audit trail by default,
+under nobody's special arrangement. And the **current** picture — the latest reading per
+`(licence, consumer, period, metric)` — is upserted into a **usage context store**, so "how close is
+this cap" and "what has this project spent this month" are a query, not a fold over the event log. The
+event is the record of what happened; the store is the answer to what is true now.
+
+**Usage is not a credential, and is recorded in the clear.** The one thing the mesh must not read is the
+sealed key ([ADR 0050](0050-model-access-is-vendor-agnostic.md), the refresh-token carve-out aside). A
+token count and a cost are not secrets; they are the operator's own operational facts, recorded openly
+so they can be queried, audited, and charged against. This is the deliberate opposite of the credential
+rule, and stating it prevents a later reader assuming usage inherits the key's secrecy and hiding it
+from the person who is paying.
+
+This is a decision and not a patch because it settles **where a model's usage lives and at what grain**,
+which every reader — a bill, a cap alarm, a per-project report — depends on, and because it closes the
+half [ADR 0050](0050-model-access-is-vendor-agnostic.md) explicitly left open, reusing the session
+([ADR 0026](0026-the-mesh-has-a-session-of-its-own.md)), the schedule
+([ADR 0053](0053-a-step-that-runs-on-a-schedule.md)), the event ([ADR 0041](0041-events-are-a-relationship.md))
+and the store ([ADR 0008](0008-a-context-owns-its-store.md)) the mesh already has rather than inventing a
+vocabulary beside them.
+
+### How each claim is checked
+
+- **A usage row is vendor-neutral and carries the raw beside it.** A unit test constructs an Anthropic
+  `utilization%` reading and a token/cost reading and asserts both render to
+  `(licence, consumer, period, metric, value)` with the vendor response preserved in the raw column;
+  a reader that answers "what did this cost" touches only the normalised columns.
+- **The two grains differ only in the consumer.** A unit test records a licence-grain row (consumer =
+  the module) and a session-grain row (consumer = a session id) for one licence and asserts both are
+  the same shape and both are returned when the licence's usage is asked for, distinguishable by
+  consumer.
+- **A poll is a scheduled run and its failure is not fatal.** The adapter's usage container declares a
+  `schedule` ([ADR 0053](0053-a-step-that-runs-on-a-schedule.md)); a host test (0053's) already proves a
+  scheduled step runs on cadence, does not gate, and logs rather than fails on a non-zero run — the poll
+  inherits this and adds nothing to check.
+- **Each reading is an event the audit trail records.** An integration check asserts a usage reading
+  emits an event that the `audit-logger` receives (it consumes `#`), so the history is present without
+  the usage module and the audit module knowing about each other beyond the event.
+- **The current picture is a query.** A store test upserts two readings for one
+  `(licence, consumer, period, metric)` and asserts the later replaces the earlier, so "what is true
+  now" is one row, while the event log keeps both.
+- **A static-key vendor with no usage reading records nothing, and that is fine.** A test resolves a
+  `static-key` model-access consumer whose adapter has no `usage` verb and asserts the licence works and
+  no usage rows or poll are required — usage is optional, holding a licence is not conditioned on it.
+- **Usage is readable in the clear; the key is not.** A test asserts a usage row is stored unsealed and
+  is returned to an ordinary query, while the licence key remains sealed and absent from the same
+  surfaces — the deliberate inversion of the credential rule.
+
+## Consequences
+
+- **A bill and a cap alarm are both queries.** "What did project X spend this month" reads the
+  session-grain rows; "how close is account Y to its cap" reads the latest licence-grain metric — both
+  from the usage store, neither a fold over events or a call to the vendor.
+- **The session becomes the unit of cost, which is what it already is.** Because a session is the
+  consumer, attributing cost needs no new identity — and the worker-granularity gap
+  ([design 14](../03-DESIGN/01-to-be/14-model-access.md)) surfaces here exactly as it does for access,
+  to be closed once, for both, when a session id is threaded through.
+- **The adapter carries the vendor's usage quirks alone.** Anthropic's `utilization%`, its transcript
+  shape, a mid-session account switch — all live in the Anthropic adapter; the mesh, the store, and
+  every reader see only rows. A second vendor adds a second adapter and no new table.
+- **Usage history is durable and tamper-evident by reuse, not by a new mechanism.** It rides the event
+  trail the mesh already keeps, so an operator who wants the whole history has it, and one who wants the
+  current number has the store — without usage owning either mechanism.
+- **The refresh-token carve-out is untouched by this.** Usage is read *from* an authenticated adapter;
+  it neither holds nor exposes the credential, so the one place the mesh reads what it stores
+  ([ADR 0050](0050-model-access-is-vendor-agnostic.md)) is not widened by recording what that credential
+  was spent on.
+
+## References
+
+- [ADR 0050](0050-model-access-is-vendor-agnostic.md) — model access is vendor-agnostic; fixes the
+  usage row shape and the `usage(licence)` adapter verb, and leaves its home open — which this closes
+- [ADR 0026](0026-the-mesh-has-a-session-of-its-own.md) — the mesh has a session of its own; a session
+  is the consumer the finer grain attributes to
+- [ADR 0053](0053-a-step-that-runs-on-a-schedule.md) — a scheduled step; the licence-grain poll is one
+- [ADR 0041](0041-events-are-a-relationship.md) — events are a relationship; a usage reading is one, and
+  the audit-logger records it
+- [ADR 0042](0042-the-shape-of-an-event-on-the-wire.md) — the shape of an event on the wire; the form a
+  usage reading takes to reach the audit trail
+- [ADR 0008](0008-a-context-owns-its-store.md) — a context owns its store; the current usage picture
+  lives in one
+- [ADR 0024](0024-model-access-is-a-provision.md) — model access is a provision; usage is a reading of
+  what that provision was used for
+- [03-DESIGN/01-to-be/14-model-access.md](../03-DESIGN/01-to-be/14-model-access.md) — the model-access
+  design; this fills its usage section and shares its open worker-naming gap
+- [03-DESIGN/01-to-be/15-the-agent-session.md](../03-DESIGN/01-to-be/15-the-agent-session.md) — the agent
+  session; the consumer the session grain is keyed to

02-DECISIONS/0055-model-access-is-answered-by-a-licence-or-a-node.md

@@ -0,0 +1,146 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-07
+deciders: jochen
+reconstructed: false
+extends: 0050-model-access-is-vendor-agnostic.md
+---
+
+# 55. Model access is answered by a licence, or by a node that hosts the model
+
+## Context
+
+**[ADR 0024](0024-model-access-is-a-provision.md) made model access a provision, and
+[ADR 0050](0050-model-access-is-vendor-agnostic.md) fixed what answers it: a record — a licence — that
+an adapter turns into a sealed vendor credential.** A consumer requires `model-access`, is put on a
+licence, and is delivered a key (a static API key, or an access token a manager refreshes). Every
+answer so far has been a credential to reach a vendor's API across the internet.
+
+**But a model need not come from a vendor. A node in the mesh can host one.** An operator with a GPU
+runs Ollama or vLLM, which serves an OpenAI-compatible API on that node. A consumer that wants that
+model does not need a vendor credential — it needs the model server's **endpoint**: the base URL and
+the model name, and a key only if the server is configured to want one. This is model access answered
+by a **node**, not by a record.
+
+**The mesh already knows how a node answers a provision — it is the ordinary provider/consumer path.**
+A provider `provides` a provision at a scope, `serves` its connection facts, and the mesh fills the
+consumer's bound facts with the provider's `at`/`port` and each served fact — exactly how a Postgres
+consumer learns where its database is. Nothing reserved `model-access` to records: a node offering
+`provides: ["model-access"]` resolves through this path, and the resolver already **prefers a local
+answer over a licence** — its own comment names the case, "a model the mesh runs itself." So the
+capability exists; what is missing is the decision to use it, and the statement of what a node-answer
+delivers and where it stops.
+
+**A node-answer and a record-answer are the same provision with two shapes of answer.** This is the
+same move [ADR 0050](0050-model-access-is-vendor-agnostic.md) already made for shapes within the
+vendor path (static-key vs refreshable-grant): one provision, more than one way it is answered. A
+consumer written against `model-access` should not care whether the model behind it is a vendor's or
+the mesh's own — it asks for model access and is given what reaches a model.
+
+## Considered Options
+
+1. **A separate provision for the local case (`local-model`, `model-endpoint`).** A node answers that;
+   `model-access` stays record-only. Rejected: it splits "where my model comes from" into two
+   provisions a consumer must choose between in its manifest, when the mesh already models a
+   record-answer and a node-answer to **one** provision. A consumer would have to know, at authoring
+   time, whether its model will be a vendor's or the mesh's — the exact coupling the provision was
+   meant to remove. It is the safer implementation (see the limitation below) but the worse interface.
+
+2. **Model access answered by either a licence or a node, under the one provision.** A consumer
+   requires `model-access`; the operator answers it with a licence (a vendor) or by assigning a
+   node that hosts a model. Adopted: one interface, and the answer is an operator's deployment choice,
+   not a consumer's authoring choice.
+
+## Decision
+
+**Model access is one provision answered two ways: by a licence (a record, turned into a sealed vendor
+credential by an adapter — [ADR 0050](0050-model-access-is-vendor-agnostic.md)) or by a node that hosts
+the model (a provider that serves an endpoint).** A consumer requires `model-access` and is delivered
+whichever the operator assigned; it does not name the kind.
+
+**A node-answer delivers an endpoint, not a credential.** The provider `provides: ["model-access"]`
+and `serves` its connection facts — the port it listens on and the model it runs — and the mesh fills
+the consumer's bound facts with the provider node's `at`, the served `port`, and the served `model`,
+the same way every provider consumer learns where its provider is. The consumer assembles a base URL
+(`http://<at>:<port>/v1`) and points an OpenAI-compatible client at it. If the local server wants a
+key, the provider mints one the ordinary way (a per-consumer secret, sealed and host-unsealed); if it
+does not — the common Ollama case — the consumer lists `model-access` under `binds` and **not** under
+`secrets`, and no key is delivered. Secret delivery and fact delivery are already independent, so a
+keyless endpoint is expressed by asking for the facts and not a secret.
+
+**A node-answer uses no adapter.** The adapter registry ([ADR 0050](0050-model-access-is-vendor-agnostic.md))
+is the vendor-credential machinery — accept-and-seal, refresh, usage. A node-hosted model has no vendor
+secret to seal; its endpoint is served, and its key (if any) is minted like any provider's. The adapter
+is consulted only for the record/vendor answer. So the vendor-agnostic decision is untouched, and the
+node-answer adds no vendor logic anywhere.
+
+**The resolver prefers a local answer.** When a node's own set answers `model-access` — a model the
+mesh runs itself — a licence for it is not consulted. This is already the resolver's behaviour and is
+made a decision here: a mesh that runs a model uses it, and a licence is the answer for a consumer that
+has no local model, not a competitor to one that does.
+
+**One limitation, stated so it is not found as a bug.** The local-preference above is exact for a
+**node-scope** provider co-located with its consumer, and for any node-answer in a mesh that holds no
+`model-access` licence. It is *not* yet exact for a **mesh-scope** model server — one node serving the
+model to others — **while a licence for `model-access` also exists in the same mesh**: the record pass
+that turns a licence into an answer keys on same-node satisfaction and would still demand the licence be
+used, double-answering. Until that pass is taught to stand down when a brokered node need already
+answers, a mesh-scope local model and a vendor licence must not both answer `model-access` in one mesh.
+A node-scope local model has no such constraint. This is named because an unstated limitation is
+indistinguishable from a bug, and costs more.
+
+This is a decision and not a patch because it settles **what may answer model access** — a question
+every model-access consumer's meaning depends on — and because it lets the mesh's own hosted models sit
+behind the same provision as the vendors', which is what makes "the mesh can run its own model" a
+deployment choice rather than a second interface to build against.
+
+### How each claim is checked
+
+- **A node answers model access without a licence, and is preferred over one.** A resolver test
+  assigns a consumer and a module that `provides: ["model-access"]` at node scope on the one node, with
+  a licence also present, and asserts no resolved need is answered by the record — the local model
+  answers and the licence is ignored. (This test exists; the decision adopts what it proves.)
+- **The consumer is delivered an endpoint, not a credential.** A mesh bed assigns a model-server
+  provider and a consumer that binds `model-access` and does not list it under `secrets`, and asserts
+  the consumer's config carries `OPENAI_BASE_URL` built from the provider's served `at`/`port`, and
+  that no key file was delivered to its secret path.
+- **The local endpoint is reachable through what the consumer was given.** The bed makes a request to
+  the base URL the consumer wrote and asserts the model server answers — the wiring, not a model's
+  output, is what is proven (the server may be a stub; a real model is not needed to prove the mesh
+  routed the consumer to it).
+- **A node-answer consults no adapter.** A node-answered `model-access` need is resolved with the
+  vendor registry never read — asserted by the absence of any vendor on a node-answered need and the
+  ordinary served-facts delivery.
+- **The scope limitation holds where stated.** The node-scope case is what the bed and the resolver
+  test exercise; the mesh-scope-plus-licence collision is recorded here and left for the resolver
+  change that reconciles the two answer passes, not worked around in a module.
+
+## Consequences
+
+- **The mesh can run its own model, and a consumer reaches it through the same `model-access` it uses
+  for a vendor.** One interface, two answers; a consumer moves between a vendor and a local model by an
+  operator reassigning its provision, not by a code change.
+- **A local model is keyless by default and keyed by the ordinary path when it must be.** Nothing new
+  is invented for the local server's credential: it either has none, or mints one the way every
+  provider does.
+- **The vendor path is untouched.** Adapters, the refresh carve-out, and usage
+  ([ADR 0054](0054-model-usage-is-recorded-at-two-grains.md)) are the record answer's business; a
+  node-answer neither uses nor changes them. A local model that exposes usage would serve it as facts,
+  not as an adapter's usage verb.
+- **The two answer passes meet in one place, and must be reconciled there.** The mesh-scope limitation
+  is the single point where a node-answer and a record-answer to the one provision can collide; it is
+  named, and its fix is a resolver change, not a per-module workaround.
+
+## References
+
+- [ADR 0024](0024-model-access-is-a-provision.md) — model access is a provision; this decides a node
+  may answer it, not only a record
+- [ADR 0050](0050-model-access-is-vendor-agnostic.md) — model access is vendor-agnostic; the record
+  answer and its adapters, which the node answer sits beside and does not use
+- [ADR 0054](0054-model-usage-is-recorded-at-two-grains.md) — model usage; a vendor's business on the
+  record path, served as facts (if at all) on the node path
+- [ADR 0008](0008-a-context-owns-its-store.md) — a context owns its store; a licence is a record in one,
+  a node-answer needs none
+- [03-DESIGN/01-to-be/14-model-access.md](../03-DESIGN/01-to-be/14-model-access.md) — the model-access
+  design, which this extends with the node answer

02-DECISIONS/0066-public-routing-is-name-agnostic.md

@@ -0,0 +1,115 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-09
+deciders: jochen
+reconstructed: false
+extends: 0007-connectivity.md
+---
+
+# 66. Public routing is name-agnostic, its names are resolved inside the mesh, and an internal authority can certify them
+
+## Context
+
+**[ADR 0007](0007-connectivity.md) and [connectivity §3](../03-DESIGN/01-to-be/08-connectivity.md)
+made a public route a grant: a workload that must be reachable requires a route, the proxy provides
+it, the consumer contributes the name it wants and the port it listens on.** What was never pinned
+is **what that name is** — and building a whole mesh in the lab showed the gap costs more than it
+looks.
+
+**The catalogue shipped each route as a full domain.** A module that needed a public name carried
+that name, in full, as a literal in its manifest. Running the same catalogue against a different
+domain — a lab standing in for production, or a second operator's mesh — meant overriding that
+literal on every routed module, per node. The mesh was, in effect, carrying a **map of names to
+services**: the one thing it should never hold, because a name is the operator's choice (one runs
+the forge at `git`, another at `code`) and the domain is the node's, and neither is the mesh's to
+know.
+
+**And a second gap surfaced the moment an internal issuer tried to certify those names.**
+[Connectivity §5](../03-DESIGN/01-to-be/08-connectivity.md) already states the issuer must be
+configurable and that the lab runs its own ACME authority. With that authority wired to the proxy,
+issuance still could not complete: the authority accepted the order and offered a challenge, then
+**could not connect to the validation target.** Nothing inside the mesh resolved the public route
+name. The mesh publishes each `<node>.internal` name into every container, but not the public names
+the proxy serves — so a validator living in the mesh had no address to reach, and a name the mesh
+cannot resolve is a name it cannot have certified.
+
+**The two are one problem.** A name the mesh can *compose* from parts it is given, and *propagate*
+to whoever needs to resolve it, is exactly a name it can also have *certified* — and the reverse:
+without the composition and the propagation, neither the routing nor the certificate is the
+operator's to move between meshes.
+
+## Considered Options
+
+**1. Keep the full domain in the manifest, override per node.** The status quo. It works, and it is
+wrong in the specific way this repository cares about: the catalogue holds a domain map, lab and
+production differ by an override on every routed module rather than one fact, and a module manifest
+names something — the public domain — that belongs to the node, not the module. An unowned name in
+the wrong place is the shape of a leak.
+
+**2. A module declares a label; the node declares its public domain; the mesh composes.** The route
+contribution carries a subdomain the operator chose, the node carries its public domain as
+node-level configuration, and the mesh joins `<label>.<public-domain>` and grants exactly that. The
+mesh interprets nothing. Lab-versus-production becomes one node setting. Chosen.
+
+**3. For certification, issue only from a publicly reachable node against a public authority.**
+This is already true for public meshes and stays true. It is not an option for a lab or an
+internal-only mesh: there is no public authority to answer, and no public reachability to validate
+against. An internal issuer is required there — and an internal issuer must be able to *validate*,
+which it cannot do unless the routed name resolves and is reachable **inside** the mesh. So the
+resolution gap is not optional to close; it is what makes an internal authority possible at all.
+
+## Decision
+
+**The mesh core holds no map of hostnames, subdomains or domains.** A route contribution carries a
+**label** (the subdomain) chosen by the module's operator. A node contributes its **public domain**
+as node-level configuration. The mesh composes `<label>.<public-domain>`, grants exactly that name,
+and never interprets what it means. One operator's forge at `git.example.tld` and another's at
+`code.other.example` are the same module with two facts supplied around it.
+
+**When the proxy is granted a name, the mesh publishes that name → the node that serves it into
+internal resolution, mesh-wide** — the same mechanism, and the same "given by the mesh, not chosen
+by a module," that already writes `<node>.internal` into every declared container
+([connectivity §2](../03-DESIGN/01-to-be/08-connectivity.md)). The mesh propagates the names it was
+told to serve. It still knows nothing about what any of them mean.
+
+**An internal authority certifies those names by the same path a public one would.** The proxy is
+pointed at whichever issuer the mesh names — a public ACME authority, or an internal one — and
+trusts that issuer's root; nothing else about issuance changes. The internal authority validates by
+reaching the routed name, which the clause above has just made resolvable inside the mesh. So the
+three are one decision: **compose the name, propagate it, certify it** — each is meaningless without
+the one before it.
+
+## Consequences
+
+- **Lab-versus-production is one node-level `public-domain` setting**, not an override on every
+  routed module. The same catalogue runs against any domain.
+- **The manifest layer needs composition it does not yet have.** Today a route name is stored as a
+  literal, with no interpolation of a node's domain into a module's label. Until that exists, the
+  composed name is produced by a per-node settings override — a stopgap that reproduces option 1's
+  per-module cost and is explicitly *not* the design.
+- **An internal issuer depends on route-name resolution.** Its challenge validates against the
+  routed name; without that name in internal resolution, issuance for it cannot complete inside the
+  mesh. The lab found this as a live failure, not a theory.
+- **Nothing about the public path changes.** A publicly reachable node issuing a public name from a
+  public authority is untouched; this widens the same shape to names and meshes that are not public.
+
+**How each is checked** — an unenforced rule is indistinguishable from a wrong one:
+
+- **Name-agnostic:** the same catalogue resolves against two different public domains by changing
+  one node setting and nothing else; and no module manifest contains a full public domain. A
+  manifest that pins an FQDN is the smell the check looks for.
+- **Resolution:** a request to a routed name, made from inside the mesh, reaches the workload that
+  serves it — and, the sharper check, issuance for that name against the internal authority
+  completes, which it cannot unless the validator resolved and reached the target.
+- **Internal authority:** a TLS handshake to a routed name verifies against the internal root and
+  nothing else, the same shape §5 already uses for internal node-to-node names.
+
+## References
+
+- [ADR 0007 — connectivity](0007-connectivity.md), which made a route a grant and named exposure,
+  resolution and certificates as one context.
+- [ADR 0009 — modules and the graph](0009-modules-and-the-graph.md), the provide/require/contribute
+  vocabulary a route and an authority both use.
+- [Connectivity design §2, §3, §5](../03-DESIGN/01-to-be/08-connectivity.md), amended alongside this
+  record.

02-DECISIONS/0067-genesis-is-a-pivot.md

@@ -0,0 +1,129 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-10
+deciders: jochen
+reconstructed: false
+extends: 0006-the-substrate-and-the-control-plane.md
+---
+
+# 67. Genesis is a pivot: a temporary control plane installs the registry that makes it permanent
+
+## Context
+
+**The mesh builds its own modules into its own registry, and there is no public registry for them
+— that is the point, not an omission.** A module is cloned from the forge, built, and published
+where the mesh can move, replace and back it up. Nothing about that arrangement wants a copy of the
+mesh's code hosted by somebody else.
+
+**Every image must be pinned by digest** ([ADR 0006](0006-the-substrate-and-the-control-plane.md)),
+and the reasoning is exactly right: a bundle is applied where no mesh exists to check anything
+against anything, so what it names must be exact.
+
+Those two sentences are individually correct and together they close a door. The digest a pin means
+is a *manifest* digest, and a manifest digest is **assigned by a registry when something is pushed
+to it**. The control plane's image is built from source and pushed nowhere, so it has no such
+digest, so it cannot be named — and a registry cannot be installed without a control plane to
+install it. That is not a pin. It is a dependency the pinning rule created by accident.
+
+**It went unnoticed because the lab hid it.** The lab raised a disposable registry, stocked it from
+a workstation, and rewrote every image reference to point at it — so the lab bootstrapped along a
+path no real machine has. A first node in the lab always worked, and a first node anywhere else had
+no path at all. Every bootstrap fault found this year was found late for the same reason: **the
+install procedure existed only as a test fixture**, and a fixture is free to invent what it needs.
+
+## Considered Options
+
+**1. Publish the mesh's own images to a public registry.** Rejected on the premise: there is no
+public registry for the mesh's modules and there is not meant to be. It would also make raising a
+mesh depend on somebody continuing to host its code, which is the dependency the whole arrangement
+exists to remove.
+
+**2. Build the control plane from source on the first machine.** Rejected. A bare machine would
+need a toolchain and a working tree — and worse, the source lives in a forge **that runs on the
+mesh**. A total rebuild would then need the mesh it is rebuilding. Acceptable for adding a node to
+a healthy mesh; useless for the case that matters.
+
+**3. Keep a disposable registry as an install step.** Rejected. It exists in no production, and
+concealing this problem is precisely what it has been doing.
+
+**4. Pivot through a temporary control plane.** Chosen.
+
+## Decision
+
+**An image may be named by the digest of its own configuration.** A bare `sha256:…` names an image
+the machine already holds — content-addressed, immutable, unforgeable, and requiring nothing to
+have served it. It satisfies what the pinning rule asks for; the rule simply never contemplated an
+image that no registry had ever seen. It is legal exactly where nothing could have served one.
+
+**Genesis is a pivot**, in this order:
+
+1. the installer **carries the control-plane image** and loads it onto the machine
+2. a **temporary** control plane is raised from it, named by that image's own digest
+3. the **registry module is installed** — its image is upstream and it is never built, which
+   [`04-ISSUES/029`](../04-ISSUES/029-the-artifact-store-cannot-be-delivered-by-the-artifact-store/00-report.md)
+   already settled: a module that provides the artifact store cannot be delivered through it
+4. the control-plane image is **pushed into the mesh's own registry**, which assigns it a manifest
+   digest — the first one it has ever had
+5. the control plane is **reinstalled as an ordinary module** pinned to that digest
+
+**The host performs the replacement, not the control plane.** Tier 0 outlives tier 2: the control
+plane composes a declaration naming the registry-pinned image, and the host applies it and recreates
+the container. Nothing is asked to replace itself while running, and the control plane is stateless
+— what it knows is in the store.
+
+**The installer is tier 0, and a separate program from the host.** Bootstrapping is by hand and
+changes the machine, which is tier 0's definition. But the host states that it *connects to nothing
+and listens on nothing*, and that claim is what makes the one thing running forever on every
+machine auditable. An installer connects to plenty. Same tier, same delivery, different program.
+
+## Consequences
+
+- **The control plane stops being a special case.** It becomes an ordinary module with an ordinary
+  image in the mesh's own registry — so the mesh can build and roll out **its own upgrades**, which
+  is what a mesh that runs itself was always reaching for.
+- **The bundle's job shrinks** to raising a temporary control plane exactly once.
+- **The source builds the installer; it does not run it.** Cloning moves to a release machine, where
+  a forge being available is an ordinary working assumption, and leaves the disaster-recovery path
+  where it very much is not.
+- **The lab's disposable registry is deleted.** The lab bootstraps by running the same program a
+  bare machine runs — the only arrangement in which the installer cannot quietly drift out of truth
+  again.
+- **Two public images remain at genesis** — the store and the broker. An air-gapped install would
+  embed those too, at a much larger artifact; that is a build variant, not a different design.
+- **A control-plane module manifest must exist**, and did not.
+- **The registry may require nothing.**
+  [`04-ISSUES/029`](../04-ISSUES/029-the-artifact-store-cannot-be-delivered-by-the-artifact-store/00-report.md)
+  states that a module providing the artifact store may not *build* artifacts, because there is
+  nowhere to put them until it runs. The pivot shows that is the narrow case of a wider rule: **it
+  may not require anything the store is needed to deliver.** Found the hard way — an unrelated
+  change gave the registry a public name and, with it, a route requirement. At genesis nothing
+  provides a route, and nothing can, because the routing stack needs images and images need the
+  store. The same cycle, re-entered through a door the existing wording did not cover.
+- **The handover is the sharp edge.** For one moment the bundle and the module both describe the
+  same container, and the host tracks what it owns. If a safe handover is not expressible with what
+  exists, the install **stops before it** and says what is missing. A machine left without a control
+  plane cannot be fixed remotely, so a partial install that halts cleanly is the better outcome.
+
+**How each is checked** — an unenforced rule is indistinguishable from a wrong one:
+
+- **Naming by its own digest:** a machine that can reach no registry at all raises a control plane.
+- **The pivot completed:** after installing, the running control plane's image is pinned by a digest
+  **the mesh's own registry assigned** — not by an image id. If it is still the image id, the pivot
+  did not happen and the mesh cannot upgrade itself.
+- **No fiction left in the lab:** the scenario declares no registry machine, and the bed bootstraps
+  through the installer rather than around it.
+- **The handover:** a machine whose control plane has been replaced still has one, and it answers.
+- **The registry requires nothing:** its manifest is resolvable on a mesh that has no other module
+  in it. A requirement added to it later is caught where it is written, rather than by a genesis
+  that cannot complete — which is how this one was found.
+
+## References
+
+- [ADR 0006 — the substrate and the control plane](0006-the-substrate-and-the-control-plane.md),
+  which pinned images by digest and named what a first node fetches.
+- [ADR 0005 — the node host](0005-the-node-host.md), which makes tier 0 the one thing installed by
+  hand and the only thing that changes a machine — the property this keeps true by shipping the
+  installer beside the host rather than inside it.
+- [`04-ISSUES/029`](../04-ISSUES/029-the-artifact-store-cannot-be-delivered-by-the-artifact-store/00-report.md),
+  the same cycle one layer down, and the rule that a registry module is named and never built.

02-DECISIONS/0068-the-lab-takes-requests.md

@@ -0,0 +1,112 @@
+---
+topic: building it
+status: proposed
+date: 2026-09-12
+deciders: jochen
+reconstructed: false
+extends: 0016-the-lab.md
+---
+
+# 68. The lab takes requests, one at a time, and runs each from its own copy
+
+## Context
+
+**The lab is exclusive hardware, and today a person holds it.** Raising a scenario takes over
+addresses and names on the workstation for as long as it stands, and only one scenario can stand
+at a time. So a run is not merely slow — it occupies the machine and the person who started it,
+who then waits rather than works.
+
+**Running it in the background against the working copy is worse than waiting.** The obvious fix
+is to start a run and carry on editing. But a run reads the working copy as it goes: binaries are
+rebuilt from it, manifests are read out of it, and the bed's own code is loaded from it. Edit
+while it runs and the result describes a state that never existed — a mixture of what was there
+when each file happened to be read. A green result obtained that way is not evidence, and a red
+one costs a day to disbelieve.
+
+**Nothing today records what was asked for.** A run is a command line in somebody's terminal. What
+commit it exercised, what it was trying to find out, and what it answered all live in scrollback,
+which is why the same question gets re-run rather than looked up.
+
+**Most of the parts already exist.** The lab writes a receipt of its last run. The mesh already
+carries messages between nodes and can notify a person. The machine already runs work on a
+schedule. What is missing is the thing in the middle.
+
+## Considered Options
+
+**1. Leave it as it is — a person drives the lab and waits.** Rejected. It is the loop
+[ADR 0010](0010-delivery.md) removed everywhere else, kept here by habit rather than by argument,
+and the cost compounds: because a run is expensive to start and blocks the person, fewer are run,
+so faults are found later and in larger batches.
+
+**2. Run in the background against the working copy.** Rejected on the reasoning above. The
+failure is silent, which is the kind this repository exists to refuse.
+
+**3. Put the lab behind the ordinary build pipeline.** Rejected for now. The pipeline builds
+artifacts and does not own a machine that can raise virtual machines; giving it one makes the
+pipeline's slowest job the lab's, and couples every push to hardware only one machine has. This
+may become right later; it is not the smallest thing that works.
+
+**4. A queue in front of the lab, and an isolated copy behind it.** Chosen.
+
+## Decision
+
+**The lab accepts requests rather than commands.** A request is recorded, queued, and answered.
+The person who made it is told when it is answered and does not wait.
+
+**A request names a bed and a commit, and nothing else.** This is the load-bearing restriction. A
+request may say *run this bed, at this version of these repositories*. It may not say what to
+install, on which machine, or with which settings — because a request that could say those things
+would be a second way of installing a mesh, and the whole reason the installer exists is that the
+lab already was one ([ADR 0067](0067-genesis-is-a-pivot.md)). The bed decides what is installed;
+the request only decides which bed and which version.
+
+**Requests are released one at a time.** The hardware admits one standing scenario, so the queue
+enforces what the hardware already requires, rather than leaving it to whoever remembers.
+
+**Every run happens in a copy the lab owns.** The lab checks the requested commit out into its own
+path and builds and runs from there. A working copy is never read by a run. This is what makes the
+queue safe to use while work continues, and without it the rest of this record is not worth
+having.
+
+**The lab is reached through tools, not only a command line.** A command line is available only
+to whoever is sitting at the machine, which is the constraint this record exists to remove. The
+lab answers three questions to anything that can reach the mesh — *what is standing now*, *what is
+queued or running*, and *what did this request answer* — and accepts a request and a cancellation.
+An agent can therefore start a run, stop attending to it, and come back; and somebody who did not
+start a run can still see it, which is the difference between a shared lab and a private one.
+
+**The restriction holds at every door.** A tool submits a bed and a commit, exactly as a command
+line does. A tool that could name a module, a node or a setting would reintroduce the second
+installer through a different entrance, and the entrance is not what made it dangerous.
+
+**Every run leaves a record that outlives the terminal**: what was asked, which commit, when it
+ran, what it answered, and where its output went. A question already answered is looked up rather
+than re-run.
+
+## Consequences
+
+Work continues while the lab runs, which is the point. A second session may edit freely, because
+nothing it edits is what the lab is reading.
+
+A request is reproducible by construction: it names a commit, so the same request can be asked
+again and compared. Today two runs of "the same thing" are only as alike as the tree happened to be.
+
+The lab gains a second copy of every repository it exercises, costing disk and needing to be kept
+from drifting into a place people edit by hand.
+
+Anything that can reach the mesh can now see what the lab is doing, including an agent working on
+something else. That is the intended gain and also the obvious hazard: a thing that is easy to ask
+is easy to ask too often, and the hardware still admits one scenario at a time.
+
+The queue becomes a thing that can fail — stuck, backed up, or lost — and a queue nobody watches
+is worse than no queue, because it absorbs requests silently.
+
+## How this is checked
+
+| Rule | Checked by |
+|---|---|
+| A run never reads a working copy | The runner is given a path it owns and no other; a run started while a working copy is deliberately dirtied produces a result matching the commit, not the edits. |
+| One scenario stands at a time | A second request submitted while one runs is observed to wait, not to raise. |
+| A request cannot say what to install | The request format admits a bed and a commit only. A request naming a module, a node or a setting is refused, and the refusal is exercised. |
+| A request is answered | Every queued request reaches a terminal state with a record. A request that vanishes is a failure of the queue, not a quiet nothing. |
+| The lab can be asked from elsewhere | What is standing is asked from a session that did not raise it, and the answer matches the machine. A lab that only answers its own caller has not left the terminal. |

02-DECISIONS/0069-a-module-is-a-repository-and-a-path.md

@@ -0,0 +1,89 @@
+---
+topic: building it
+status: accepted
+date: 2026-09-12
+deciders: jochen
+reconstructed: false
+extends: 0009-modules-and-the-graph.md
+---
+
+# 69. A module is a repository and a path within it
+
+## Context
+
+**The builder clones one repository and reads `module.json` at its root.** The to-be design says
+so in as many words — *"one file at the root"* — and the code implements it: clone, read the root
+manifest, build what it declares.
+
+**Nothing that exists is shaped that way.** The catalogue holds sixty-seven modules, each in its
+own directory, and has no manifest at its root. None of the five code repositories has one either.
+So today the builder cannot be asked to build any module that exists: pointed at the catalogue it
+finds no manifest, and pointed at a module's source it finds no manifest.
+
+**The system being replaced already works the other way**, and has for years: a monorepo with one
+directory per piece of software, and the coordinator builds a module from a repository and a path
+inside it. The root-only assumption is not a simplification of that — it is a different model that
+was never reconciled with it.
+
+**And it splits what a build needs into two places.** The control plane's manifest sits in the
+catalogue; the source it describes sits in the control plane's own repository. A build must read
+one tree, so under the root-only model neither location can be built from.
+
+## Considered Options
+
+**1. One repository per module.** Rejected. Sixty-seven repositories for sixty-seven modules, most
+of which are a single manifest naming a public image, and every one needing its own creation,
+permissions and lifecycle. It also contradicts [ADR 0015](0015-applications-live-in-their-own-repository.md),
+which put *applications* in their own repositories precisely because modules do not need one.
+
+**2. Keep manifests in the catalogue and source elsewhere, and have a build fetch both.** Rejected.
+A build would clone two trees whose versions can disagree, so "what commit is this module?" stops
+having one answer — and that question is the whole basis of knowing when to rebuild.
+
+**3. A module is a repository and a path within it.** Chosen. It is what the current system does,
+what the catalogue already looks like, and it keeps a module's description beside the thing it
+describes.
+
+## Decision
+
+**A module is named by a repository and a path within it.** The path holds `module.json`, and
+everything that manifest declares is produced from that path. A module whose path is the root is
+the ordinary case of this, not a separate one.
+
+**A module's manifest lives beside its source.** Where a module has code, its directory holds both,
+so one commit answers "what is this module, and what is it made of". Where a module has no source —
+a manifest naming a public image — the directory holds only the manifest, and there is nothing to
+build.
+
+**This moves the core modules.** The control plane and the builder are built from the control
+plane's repository, so their manifests belong in that repository at their own paths, not in the
+catalogue. The catalogue keeps the modules whose source it holds, and the modules that are only a
+manifest.
+
+**One commit, one module version.** Because a module is one path in one repository, the commit that
+built it identifies it exactly, and "the source has moved ahead of what the mesh holds" stays a
+question with a yes or no answer.
+
+## Consequences
+
+The builder gains a path alongside the repository and the ref. A build is `repository, path, ref`,
+and the manifest it returns is the module the mesh records.
+
+The catalogue stops being the place every manifest lives, and becomes the place manifests live
+*when their module has no other home*. That is a smaller claim than it sounds: most of the
+sixty-seven stay exactly where they are.
+
+Two repositories change shape — the control plane's gains manifests for the modules built from it.
+Nothing else moves.
+
+A repository can hold modules that are built and modules that are not, and no rule distinguishes
+them beyond whether their manifest declares anything to build.
+
+## How this is checked
+
+| Rule | Checked by |
+|---|---|
+| A module is buildable from its repository and path | The builder is asked for a module by repository and path, and returns a manifest whose artifacts are pinned to digests the mesh's registry assigned. |
+| A manifest sits beside what it describes | A module declaring something to build, whose path holds no source to build it from, is refused at build time rather than producing an empty result. |
+| One commit identifies one module | Two builds of the same repository, path and commit produce the same digests. |
+| The core modules are built like any other | The control plane is rebuilt from its own repository and path, and the running mesh is upgraded to it — the same path an ordinary module takes. |

02-DECISIONS/0070-the-catalogue-owns-the-module-graph.md

@@ -0,0 +1,107 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-12
+deciders: jochen
+reconstructed: false
+extends: 0067-genesis-is-a-pivot.md
+---
+
+# 70. The catalogue owns the module graph, and genesis builds rather than carries
+
+## Context
+
+**The module graph has no owner.** What modules exist, what each requires and provides, what each
+claims, what each is made of — all of it lives inside the control plane because that is where it
+was first written, not because anything decided it belonged there.
+
+**The control plane's own test says it does not belong there.**
+[`06-the-control-plane`](../03-DESIGN/01-to-be/06-the-controller.md) defines the tier as
+*everything that needs to know about more than one node*, and states the corollary plainly:
+anything a single machine could answer alone is not the control plane's. What a module is, and
+what it needs, requires no knowledge of any node whatsoever.
+
+**And nothing can query it.** The graph is the thing that answers *what must be rebuilt when this
+changes*, *what would break if this were removed*, and *what can be installed here* — and today it
+is reachable only as control-plane internals. [ADR 0009](0009-modules-and-the-graph.md) already
+decided that a build edge is derived, that an artifact is stale when anything it was built against
+moved, and that the rebuild set is therefore computable. None of that has anywhere to live.
+
+**Genesis currently carries an image, and cannot produce a builder at all.**
+[ADR 0067](0067-genesis-is-a-pivot.md) has the installer carry the control plane's image. That
+works, but it leaves the builder with no route onto a fresh mesh — it cannot be fetched from the
+public internet, because the mesh builds it, and the installer carries one image only. So a raised
+mesh cannot build anything, including the modules it is made of.
+
+## Decision
+
+**The catalogue is a core module, beside the control plane and the builder, and it owns the module
+graph.** Modules, what they require and provide, what they claim, their dependencies and their
+build edges, assignments and configuration — the catalogue holds them and serves tools over them:
+install a module, query what is available and what it needs, query the graph, query and update
+settings.
+
+**It is one per mesh**, expressed the way the control plane already expresses it — a claim scoped
+to the mesh, not a new mechanism.
+
+**The control plane consumes it.** Resolving what a module requires into an actual binding, and
+composing what a machine should be, both need to know what modules are. So the dependency runs
+from the control plane to the catalogue, which is the opposite of what the tiers suggest and is
+therefore written down here rather than left to be inferred.
+
+**Genesis builds the core modules rather than carrying them.** The installer ships an *init
+builder* — the one thing carried — which is started, clones the source, and builds the control
+plane, the catalogue and the builder. The installer then raises a temporary control plane, which
+installs the catalogue, registers the permanent control plane and the builder, and assigns all
+three to the first machine. The temporary control plane stops; the installer verifies that the
+permanent one answers and can query the graph. The builder then sees a catalogue in its initial
+state and builds the core modules into it.
+
+**So exactly one thing is carried, and it is a builder rather than a result.** That is the
+difference from [ADR 0067](0067-genesis-is-a-pivot.md), which carried the control plane's image:
+carrying a builder produces every core module on the machine, including the builder itself, so
+there is no component left without a route.
+
+## Consequences
+
+The catalogue joins the small set of things that cannot arrive through the ordinary path, because
+it cannot be installed by something that needs it in order to install anything. It arrives the same
+way everything else does under this record — built by the init builder before the mesh can install
+anything — so the set is answered by one mechanism rather than three special cases.
+
+The rebuild fan-out gains a home. *What was this built against* and *what must rebuild now* are
+questions about the graph, and the graph now has an owner to hold the edges and answer them.
+
+The catalogue holds state, so it owns a store in the substrate's database, the same way the control
+plane's contexts do. That is the mesh's own store and not the `postgres` module, which is a
+provider other modules consume.
+
+A mesh without a catalogue cannot resolve anything, where previously it merely lacked an interface.
+That is the cost of ownership over surfacing, and it is deliberate: one owner beats two copies.
+
+## Open, and to be settled before this is built
+
+**Where the init builder clones from.** [ADR 0067](0067-genesis-is-a-pivot.md) rejected building
+from source at genesis partly because the source lives in a forge that runs on the mesh, so a
+total rebuild would need the mesh it is rebuilding. Carrying a builder answers the toolchain half
+of that objection and not this half. Genesis must therefore name a source that exists before the
+mesh does.
+
+**What the init builder publishes into.** Building produces artifacts that must be pinned by a
+digest a registry assigned, and today the registry is installed after the machine has joined. If
+the core modules are built first, the registry has to exist first, so the substrate's order needs
+restating rather than assumed.
+
+**Where the line falls between the catalogue and the control plane.** Claims and assignments need
+to know about every node, which by the control plane's own test is its work. Whether the catalogue
+holds them and asks, or the control plane holds them and the catalogue surfaces them, is not
+settled here.
+
+## How this is checked
+
+| Rule | Checked by |
+|---|---|
+| The catalogue owns the graph | The control plane answers *what does this module require* by asking the catalogue, and a mesh whose catalogue is stopped cannot resolve — observed, not assumed. |
+| One per mesh | A second catalogue assigned anywhere in the mesh is refused by the claim, and the refusal is exercised. |
+| Genesis builds rather than carries | The installer carries exactly one artifact, and after installing, every core module is pinned to a digest the mesh's own registry assigned. |
+| The builder has a route | A mesh raised by the installer, with no hand-placed image, can build a module. |

02-DECISIONS/0071-where-genesis-gets-its-source.md

@@ -0,0 +1,80 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-12
+deciders: jochen
+reconstructed: false
+extends: 0070-the-catalogue-owns-the-module-graph.md
+---
+
+# 71. Genesis clones from a mesh, and checks what it got
+
+## Context
+
+**[ADR 0070](0070-the-catalogue-owns-the-module-graph.md) has the init builder clone the source,
+and does not say from where.** [ADR 0067](0067-genesis-is-a-pivot.md) had already rejected building
+at genesis partly for that reason: the forge holding the source runs *on* the mesh, so a total
+rebuild would need the mesh it is rebuilding.
+
+That objection is real but narrower than it reads. It only binds when the mesh being raised and the
+mesh holding the source are the same one, which is true exactly once.
+
+## Decision
+
+**Genesis clones from a mesh's forge, reached by name.** Any mesh that holds the source can serve
+it. The first mesh is not structurally special — it is simply the only one that existed when there
+was nothing else to clone from.
+
+**If the mesh serving the source is lost, the name moves to another mesh that holds a copy.**
+Recovery is a name pointing somewhere else, not a backup being restored. This is what makes the
+source's survival a property of there being more than one mesh, rather than a property of somebody
+having remembered to take a copy. A mesh that has installed from that name holds the source
+afterwards, so every installation adds a place the name could point.
+
+**Genesis names a commit and checks what it got.** It does not clone whatever a branch happens to
+point at. The forge a mesh installs from is the trust anchor for everything that mesh will ever
+run, and a branch is a moving target that somebody else controls.
+
+*This is not hypothetical.* On 2026-09-11 the forge that would serve this role was running a
+cryptominer, and its git operations were being tampered with in flight — output injected into the
+protocol stream by a hook that fired on every fetch. Nothing was altered: the repositories were
+verified against local copies and found byte-identical. But a mesh installing from that name during
+those hours had no way to establish that for itself, and would have had none.
+
+## Consequences
+
+The init builder needs a name it can resolve and a commit it can verify, and nothing else. It does
+not need to know which mesh answers.
+
+Whoever operates the mesh that name points at carries a responsibility to everyone installing from
+it, and should know that. It is not merely a convenience host.
+
+A mesh that cannot reach any forge cannot be raised. That is a real limit and it is accepted: the
+alternative is carrying the whole source in the installer, which makes the installer a release
+artifact that goes stale rather than a program that fetches what it was told to.
+
+## Open — what relationship a mesh keeps afterwards
+
+**Not decided, and named here so it is not decided by accident** by whoever writes the init
+builder. Two shapes, and they are meaningfully different:
+
+**A snapshot, and then independence.** A mesh installs once, mirrors the source into its own forge,
+and has no upstream afterwards. It is fully self-hosted, in the sense that nothing it needs lives
+anywhere else. Updates are then something an operator does deliberately, by pulling changes in —
+tooling for which is possible and is not a priority.
+
+**A continuing upstream for core modules**, the way a distribution serves packages and a separate
+collection serves everything else. A mesh keeps looking at the origin for the modules that make a
+mesh a mesh, and holds its own for the rest.
+
+The first is more obviously aligned with the rest of this design, which is arranged so nothing a
+mesh needs depends on somebody else continuing to host it. The second is more convenient and makes
+a security problem in one forge everybody's problem. Neither is chosen here.
+
+## How this is checked
+
+| Rule | Checked by |
+|---|---|
+| Genesis needs only a name and a commit | A mesh is raised with the name pointed at a different mesh than the last time, and the result is identical. |
+| What was cloned is what was asked for | Genesis is pointed at a commit and refuses a forge serving different content under it, rather than building what it received. |
+| Losing the serving mesh is survivable | The name is repointed at a mesh that installed from it earlier, and a raise succeeds. |

02-DECISIONS/0072-two-graphs-and-the-build-chain.md

@@ -0,0 +1,86 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-12
+deciders: jochen
+reconstructed: false
+extends: 0070-the-catalogue-owns-the-module-graph.md
+---
+
+# 72. Two graphs, and a build chain that orders itself
+
+## Context
+
+[ADR 0070](0070-the-catalogue-owns-the-module-graph.md) gave the catalogue the module graph and
+said the control plane consumes it — that resolving a requirement and composing what a machine
+should be *"both need to know what modules are, so the dependency runs from the control plane to
+the catalogue."*
+
+**That paragraph is wrong, and this record corrects it.** It was written before the two graphs had
+been told apart, and it creates a dependency that does not need to exist: a catalogue that is down
+would leave the control plane unable to compose the declaration that would repair it.
+
+## Decision
+
+**There are two graphs, with different owners, and they meet only when something is installed.**
+
+| Graph | Owner | What it links | Answers |
+|---|---|---|---|
+| the module graph | the **catalogue** | module-versions to each other | what was this built against · what must rebuild now · what does this need |
+| the runtime graph | the **control plane** | module-versions to nodes | what runs where · who consumes this provision · what breaks if this machine goes |
+
+The catalogue does not know nodes exist. The control plane holds module-versions and nodes, along
+with capabilities and claims, because deciding whether a machine qualifies needs every node.
+
+**So the control plane never asks the catalogue anything.** It holds what it needs to compose a
+declaration. A catalogue that is down stops new installs and stops the rebuild fan-out, and does not
+touch anything already running or the ability to repair it.
+
+**Everything between them travels as events over the broker**, like all module-to-module
+communication. The builder finishes and announces that a module was built. The catalogue registers
+it, places it among what it depends on, and announces that a module was upgraded. The control plane
+reacts to *that*, not to build output — a semantic fact rather than an artifact.
+
+**Nothing is lost if a receiver is down.** A consuming module's queue is durable with a dead-letter
+exchange, declared by the mesh rather than by the module, so an event waits for a consumer that is
+not there.
+
+**Build order is not computed. It emerges from the chain.** The builder never consults the graph: it
+builds what it is asked for, one at a time. The catalogue asks for the next build after the previous
+registration, so *"do not start this until that is registered"* holds by construction rather than by
+a schedule somebody maintains.
+
+**The catalogue's rule is a condition, not an order:** ask for a module to be rebuilt once everything
+it was built against is current. That handles a chain and a diamond with one rule, where an
+order-based approach needs to know the shape in advance.
+
+**Two refusals belong to the catalogue.** A cycle, because the chain would never settle. And a
+rebuild whose artifacts are identical to what it replaced, which is not an upgrade and must not be
+announced as one — or a single change ripples outward forever through modules that did not change.
+
+**Genesis does none of this.** The init builder has a fixed, short list — control plane, catalogue,
+builder — in a written order, because there is no catalogue yet to ask.
+
+## Consequences
+
+The builder stays simple, and independent of the catalogue. That is what makes genesis possible at
+all: the thing that builds the catalogue cannot require the catalogue.
+
+The two sides can be briefly out of step — the catalogue may hold a module-version a moment before
+the control plane knows of it. Assigning in that instant fails, and should say why rather than
+report that no such module exists.
+
+A module's declared events stop being documentation and become its permissions: an account is
+scoped from what a module emits and consumes, so a builder that announces what it built is granted
+what it needs by the ordinary mechanism rather than by a special case.
+
+## Open
+
+**Whether an upgrade is applied or merely noticed.** Today the system this replaces deploys
+automatically, and that is a defensible default for core modules on a mesh its operator runs. But
+the design as it stands does the opposite: it records that the source moved ahead, makes it visible,
+and waits to be told. This must become a setting with a chosen default rather than inherited
+behaviour — and the choice matters most on the day a bad commit reaches something that carries mail.
+
+**Whether a module assigned to several machines upgrades on all of them at once.** Doing so makes
+one bad commit simultaneous everywhere. Doing one machine and pausing turns it into one casualty.

02-DECISIONS/0073-the-installer-carries-a-builder.md

@@ -0,0 +1,94 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-13
+deciders: jochen
+reconstructed: false
+extends: 0070-the-catalogue-owns-the-module-graph.md
+---
+
+# 73. The installer carries a builder, and the registry stays where it is
+
+## Context
+
+[ADR 0070](0070-the-catalogue-owns-the-module-graph.md) decided that genesis builds rather than
+carries, and [ADR 0071](0071-where-genesis-gets-its-source.md) settled where it clones from. Two
+questions were left open, and the design record names them as the one gap that stops a fresh mesh
+from being able to produce anything at all: **how the builder arrives**, and **what it publishes
+into**.
+
+Today the installer carries the control plane's image inside itself. That works, and it is why
+genesis needs no registry: nothing is ever fetched, because the one image that matters is already
+present. The cost is that the mesh which results holds an artifact it did not make, cannot rebuild,
+and knows nothing about — no version, no source, no edges. That is the same shape as the fault
+[issue 044](../04-ISSUES/044-the-runtime-every-module-builds-on-cannot-be-built-by-the-mesh/00-report.md)
+recorded for the shared runtime, and fixing it there while shipping it here on every new mesh would
+be a strange place to stop.
+
+## Decision
+
+**The installer carries a builder, and nothing else.** One artifact, not a growing set. It clones
+the source at a named commit, checks what it got ([ADR 0071](0071-where-genesis-gets-its-source.md)),
+and produces the control plane from the same repository and path that any later rebuild of it would
+use. What raises the mesh is therefore the same thing that will maintain it, and there is no second
+mechanism kept in step with the first.
+
+**The registry does not move, and the argument for moving it does not survive being made.**
+
+It was put this way: a produced image has to be put somewhere before anything can fetch it, so the
+registry must now precede the control plane, and
+[ADR 0033](0033-the-substrate-is-a-store-and-a-broker.md)'s answer to that question has to flip.
+
+It does not, because the premise is false. **The thing that builds the image and the machine that
+runs it are the same machine.** A built image is already in that machine's container runtime, and
+the temporary control plane names it exactly as it names a carried one — by the digest of its own
+configuration, a local identity that requires nothing to have served it. Building changes where the
+bytes came from. It does not change where they are.
+
+| | is it substrate? | must it precede the control plane? |
+|---|---|---|
+| the store | yes | yes — there is nowhere else to put the control plane's state |
+| the broker | yes | yes — the control plane reaches a machine only over it |
+| the image registry | yes — it cannot grant itself a repository | **still no** — the first machine neither fetches the control plane nor needs to, whether the image was carried in or made here |
+
+So the registry stays where [ADR 0033](0033-the-substrate-is-a-store-and-a-broker.md) put it:
+substrate by role, ordinary by delivery, installed by the temporary control plane as its first act.
+The bundle carries two services and a control plane, as it did. **What publishes into the registry
+is unchanged too** — the existing step that pushes the control plane's image into it, which is the
+moment that image first receives a digest assigned by something other than itself. It now pushes
+something this mesh built rather than something it was handed.
+
+## Consequences
+
+**Genesis gains one step and changes no others.** A build happens before the image is loaded. The
+pivot described in [ADR 0067](0067-genesis-is-a-pivot.md) survives exactly as written, because the
+step it pivots on never cared where the image came from.
+
+**A fresh mesh can produce from the moment it exists.** The builder is present before the control
+plane is, so the core modules, the catalogue and the builder's own module can be built in the
+ordinary way rather than waiting for somebody to carry them in. The paragraphs in
+[`17-raising-a-mesh`](../03-DESIGN/01-to-be/17-raising-a-mesh.md) that describe this were describing
+something that could not start; they can start now.
+
+**Genesis needs more of the outside world.** Carrying an image needed nothing but the installer.
+Building one needs the source, and whatever the build itself reaches for. This is a real cost and
+is not waved away: it makes genesis fail in more ways, all of them at a step that says what it was
+doing. It is accepted because the alternative is a mesh that cannot rebuild its own control plane,
+which fails in exactly one way, silently, later, and for ever.
+
+**A pre-built bundle remains possible and is not this.** Nothing here forbids delivering artifacts
+rather than building them; it fixes where they may come from. A bundle of pre-built core modules is
+an **export of a mesh that built them**, carrying what the catalogue knows about each alongside the
+artifact itself — so that loading one leaves the graph in the state building would have left it. A
+bundle that carries images without that is the thing this decision rejects, whoever ships it.
+
+## What this does not decide
+
+**Whether the builder's own module is carried or built.** It builds everything else; what installs
+*it* as an ordinary module afterwards, so that it too can be upgraded, is the same closed-list
+question [`12-a-module-repository`](../03-DESIGN/01-to-be/12-a-module-repository.md) already holds,
+and is unchanged by this.
+
+**How a machine authenticates to a registry that asks it to.** Genesis raises its own and reaches it
+over the loopback, so this remains a joining problem
+([issue 042](../04-ISSUES/042-nothing-gives-a-node-an-account-for-a-registry/00-report.md)).

02-DECISIONS/0074-the-wire-is-specified-not-the-types.md

@@ -0,0 +1,172 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-15
+deciders: jochen
+reconstructed: false
+extends: 0039-what-the-sdk-holds-and-refuses.md
+---
+
+# 74. The mesh defines a module protocol; an SDK is an implementation of it
+
+## Context
+
+[ADR 0039](0039-what-the-sdk-holds-and-refuses.md) says what the SDK holds: the tool-serving
+harness, the messaging and event framework, the contracts, and core primitives. It settles what
+belongs in *an* SDK. It does not say what happens when there is more than one.
+
+There is already more than one. **The contracts are expressed twice** — as Go types in the control
+plane and the host, and as TypeScript types in the SDK — and nobody has felt it because both live
+in one repository and one head.
+
+**A correction, made after inspecting the wire rather than the types** (2026-09-16). This record
+first claimed the two implementations already disagreed — `resource` vs `Provision`, `consumer`
+meaning the module in one and the node in the other, headers declared on one side and emitted by
+neither. **On inspection the live wire agrees**, and the claim was wrong:
+
+- The grant types that disagreed (`Grant`, `Interface`, `Credential` in the SDK's `contracts`)
+  were **dead** — exported and imported by nothing. The live provisioning wire is the contributions
+  file, whose shape (`as`, `secret`, `node`, `at`, `values`) is the same on both sides. Those dead
+  types have been removed.
+- The envelope agrees too: Go emits all five required headers, and `x-causation-id`/`x-schema` are
+  **optional** — the SDK sets them when a handler has a causation or a schema, and a bare event
+  carrying neither is correct, not a drift.
+
+So the danger was never live disagreement. It was **dead types that contradicted the live wire**,
+which read as the contract and were not — and are exactly what led this record to assert a drift
+that inspection did not find. That is a sharper reason for the decision below, not a weaker one: a
+type is only as good as its being the wire, and the way to guarantee that is to specify the wire and
+check implementations against it, rather than to trust a hand-kept type to still describe it.
+
+A failure of this kind does not announce itself. Two implementations that disagree about an
+envelope do not fail to compile — they ignore each other's messages, and a mesh where a module
+stops reacting looks exactly like a mesh where nothing happened.
+
+## The question this settles
+
+A module may be written in any language the mesh can build
+([`18-building-a-module`](../03-DESIGN/01-to-be/18-building-a-module.md)). Every language needs an
+SDK. What is an SDK *of*?
+
+Two answers were available, and the obvious one is wrong.
+
+**Shared types, generated.** Write the shapes once — a schema, an IDL — and generate Go, TypeScript,
+Rust. It is the familiar answer and it solves the smaller half of the problem. The shapes are not
+where the difficulty is.
+
+**A specified wire, with a conformance suite.** The shapes are a consequence; what an SDK must get
+right is *behaviour*.
+
+## Decision
+
+**The mesh defines a module protocol. An SDK is an implementation of that protocol in one
+language, and nothing more.**
+
+That is the whole of what an SDK is. Not a library a language happens to have, not a convenience
+layer, not a place for helpers to accumulate — an implementation of a specified protocol, finished
+when it implements it and correct when it agrees with every other implementation.
+
+### The protocol is split per capability
+
+**A module does not use all of it, so an SDK need not implement all of it.** A module that only
+consumes events uses the event capability. One that serves tools uses the tool capability. A
+provider uses provisioning. Nothing about consuming an event requires knowing how a grant is
+answered.
+
+So the protocol is a floor plus capabilities:
+
+| part | what it covers | who needs it |
+|---|---|---|
+| **connection** — the floor | reading the sealed credential, pinning the certificate fingerprint, taking identity from the credential rather than the environment | everything |
+| **events** | the envelope and its headers, the durable per-consumer queue, binding, at-least-once with dedup on `x-event-id` | a module that emits or consumes |
+| **tools** | registration, the shared durable `serve.<key>` queue, request and reply | a module with a surface |
+| **provisioning** | a grant in, a credential out, and what each carries | a module that provides something |
+
+**This is the same shape the host already has.** A host declares which resource kinds it can apply,
+and a partial host — one that can write files and run things but not manage users or containers —
+is a real thing rather than a broken one ([ADR 0005](0005-the-node-host.md)). An SDK that implements
+the floor and events is exactly as legitimate, and a module written against it is a module that
+does events.
+
+**So a language arrives in pieces rather than all at once.** A Rust SDK implementing connection and
+events is useful the day it exists; tools and provisioning follow when something needs them. The
+alternative — a language is unsupported until it is entirely supported — is what makes adding one a
+project rather than a contribution.
+
+**And what a language can be used for is then a fact the mesh can state**, rather than something an
+author discovers by writing a module that cannot be built: the toolchain list says which languages
+exist, and the conformance results say what each can do.
+
+### What the specification covers
+
+Per capability, what two implementations can disagree about:
+
+- **the exchanges and queues** — which exchanges exist, that a consumer's queue is durable and
+  named `<node>.<module>.events`, that a tool is served from a shared durable `serve.<key>`
+- **the envelope** — every header, which are required, what an unknown `x-` header means, and that
+  ignoring one is correct rather than lax
+- **identity** — that a module's node and module name come from its sealed credential and not from
+  its environment, so what it emits matches what the mesh authorised
+- **delivery** — at-least-once, and that dedup is on `x-event-id`, which only the emitter can make
+- **the credential** — the sealed document's fields, and that a connection pins a certificate
+  fingerprint rather than trusting an authority
+- **provisioning** — a grant in, a credential out, and what each carries
+- **the vocabulary** — that `consumer` is one thing, named once
+
+### Conformance is per capability
+
+**A suite per part, and an SDK claims the parts it passes.** A monolithic pass/fail would make a
+partial implementation indistinguishable from a broken one, which is the distinction this is built
+on.
+
+**And the suite is executable, not prose.** A specification nobody can run is a document two
+implementations drift from while both believe they conform. Conformance is a set of fixtures — an
+emitted event, a served tool call, a grant and its answer — that every SDK must produce and consume
+byte-for-byte.
+
+**The existing two implementations are the first two to be made to pass it.** Not a future language:
+the drift above is present, and a suite that only new SDKs must satisfy would leave the disagreement
+that already exists in place while certifying everything added afterwards against it.
+
+## Why not generated types
+
+Generation makes the shapes agree and leaves everything that matters unspecified. Two SDKs
+generated from one schema can still name their queues differently, take identity from the
+environment, dedup on the wrong field, or omit a header the other requires — and every one of those
+is a mesh that runs and quietly does not work.
+
+It also makes the contract into whatever the generator supports, which is a decision nobody made
+about a boundary everything else depends on.
+
+**The shapes are worth generating once the wire is specified.** That is a convenience, and it comes
+second.
+
+## Consequences
+
+**A language is a commitment, and now a divisible one.** Adding one means implementing the protocol
+and passing the suites for the parts it claims. That is more work than transliterating types, and
+it is the work that was always there — the difference is that it can be finished, and finished in
+pieces, rather than believed.
+
+**Versioning becomes possible.** `x-schema` exists for it and is never written. A specified envelope
+with a version on the body is what lets a mesh hold a module built against an older SDK, which is
+the ordinary state of any mesh that has been running for a while.
+
+**The two current implementations agree on the live wire** — inspection showed it. What was wrong
+was a set of dead types beside the wire, now removed. The suite's job here is therefore prevention:
+to keep that agreement true as the wire changes, and to hold a new language's SDK to it, rather than
+to repair a break that exists today.
+
+**This does not make the mesh polyglot by itself**, and should not be reported as though it does. It
+makes polyglot possible to do correctly. A Rust SDK is still a Rust SDK.
+
+## How this is checked
+
+| Rule | Checked by |
+|---|---|
+| One vocabulary | A word means one thing across implementations, checked by the fixtures using it. |
+| The wire is what is specified | Both existing SDKs run the conformance suite in their own test suites, and a change to one that breaks a fixture fails there rather than in a mesh. |
+| A new SDK is a passing SDK | A language is not listed as buildable for a capability until its SDK passes that capability's suite; the toolchain list and the conformance results name the same set. |
+| A partial SDK is a real thing | An SDK implementing the floor and one capability passes, is listed for that capability, and a module using another is refused with the reason — rather than failing at runtime in a language nobody said was finished. |
+| An unknown header is ignored | A fixture carries one, and every implementation accepts it. |
+| Identity comes from the credential | A fixture sets an environment that disagrees with the credential, and the emitted event carries the credential's. |

02-DECISIONS/0075-two-stores-and-which-provides-what.md

@@ -0,0 +1,129 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-15
+deciders: jochen
+reconstructed: false
+extends: 0014-no-npm-workspace.md
+---
+
+# 75. An artifact store is a provision; a package registry is a different one
+
+## Context
+
+Two questions have been circling, and they turn out to be one question asked twice.
+
+**"Should gitea be the mesh's registry?"** It serves OCI images and a dozen package ecosystems, it
+is already needed — genesis clones from one — and the mesh's own registry has neither
+authentication ([issue 042](../04-ISSUES/042-nothing-gives-a-node-an-account-for-a-registry/00-report.md))
+nor a transport a runtime will accept over a network
+([issue 048](../04-ISSUES/048-nothing-makes-a-machine-trust-the-mesh-registry/00-report.md)). Gitea
+has both.
+
+**"Where does the SDK come from?"** [ADR 0014](0014-no-npm-workspace.md) already answers it — each
+module consumes its dependencies, the mesh's own shared library included, *from the private
+registry* — and nothing installs one, so today it comes from a git URL, which is
+[issue 053](../04-ISSUES/053-the-sdk-is-pinned-twice-and-the-two-disagree/00-report.md).
+
+**The framing that dissolves both:** `artifact-store` is already a provision, and `registry`
+already provides it. So "should gitea be the registry" is not a question about replacing a
+component. It is a question about **a second provider of an existing provision** — which this mesh
+has a mechanism for, and uses for certificate authorities and VPNs already.
+
+## Decision
+
+**Two provisions, because they are two jobs.**
+
+| provision | is | for |
+|---|---|---|
+| `artifact-store` | content-addressed blobs, pinned by digest, no versions, no ranges | what the **mesh** delivers to **machines** |
+| `package-registry` | an ecosystem's own registry — npm, cargo, PyPI, Go | what **code** resolves when it is compiled |
+
+They are not the same store with different clients. One is addressed by digest and immutable by
+construction; the other is addressed by name and version, and resolves ranges. Conflating them is
+how a mesh that pins everything ends up rebuilding one commit into two different things.
+
+**`registry` remains the provider genesis installs.** Not because it is better, but because of what
+it is: a directory and one container, no database, no control plane, installable at step 8 of an
+install where neither exists yet. Gitea needs a store and provisioning, which means a control plane,
+which means the pivot has already happened — and the pivot needs somewhere to publish to.
+
+**Gitea also provides `artifact-store`, and a mesh may choose it.** Two providers of one provision
+is a thing the mesh understands: it refuses, names both, and choosing is assigning the one you want.
+
+**And it does not claim `the-artifact-store`.** That claim is node-scoped, so a module holding it
+cannot share a machine with another that does — and a machine running gitea for git and packages
+*alongside* a registry serving artifacts is an ordinary arrangement, not a conflict. They are
+different ports doing different jobs.
+
+The exclusivity that matters is mesh-wide and is already expressed: `provides` at mesh scope means
+two providers are two answers, and the resolver refuses until one is assigned. Forbidding
+co-residence adds nothing to that and forbids something reasonable. **Whether `registry` should
+still hold that claim is left open here** — it may be protecting something about the port or the
+data directory that is not written down, and removing a claim is not a thing to do from the outside
+of a manifest.
+A mesh that assigns gitea gets authentication and TLS for its artifacts — which is to say, **issues
+042 and 048 are answered by choosing a provider that already solved them**, rather than by
+reimplementing accounts and certificates in a registry that has none.
+
+**Gitea provides `package-registry`.** That is ADR 0014's private registry, and it is one service
+rather than one per ecosystem. `verdaccio` may provide it too, for npm alone, and is then a choice
+somebody makes rather than the answer.
+
+**The registry is not removed at the end of installing.** A mesh that never runs gitea still has an
+artifact store. Retiring it is a migration a mesh performs, not a step an installation ends with.
+
+## Why not simply gitea, from the start
+
+Because genesis would need a control plane before the thing that stores the control plane's image,
+and that is circular rather than merely awkward. It would also make one of the three things the
+build loop cannot produce for itself into a stateful application with a database — the pivot is the
+hardest part of this design already.
+
+And it puts every artifact in the service that is also the trust anchor for everything the mesh will
+ever run ([ADR 0071](0071-where-genesis-gets-its-source.md)), which records that forge serving a
+cryptominer with tampered git operations. Two blast radii are better than one.
+
+## Moving from one provider to the other is a designed act
+
+**Not a removal.** Every image a machine runs is pinned to a digest at a named store, the control
+plane's own included. Changing the provider means:
+
+1. gitea installed, reachable, and holding an account the builder may publish with
+2. every artifact mirrored
+3. every declaration re-pinned, the control plane's **last**, because it is what performs the others
+4. **every machine verified to have converged and to be able to pull from the new store**
+5. only then the old provider unassigned, and its volume kept ([ADR 0030](0030-data-outlives-the-mesh-that-declared-it.md))
+
+Step 4 is the one that is easy to skip and the only thing between this and a mesh that cannot
+restart its own control plane. A machine that reboots mid-migration pulls from a store that no
+longer exists, and a local image cache hides that until exactly the moment it matters.
+
+## Consequences
+
+**The bootstrap is unchanged**, which is the point of keeping the small provider.
+
+**042 and 048 gain a second answer.** They can be fixed in the registry, or dissolved by choosing a
+provider that already has accounts and TLS. The second is less work and more service.
+
+**ADR 0014 becomes satisfiable.** There is a provision for the private registry, something that
+provides it, and a module may depend on it — so the SDK can be published and consumed rather than
+cloned, and issue 053 has somewhere to go.
+
+**A mesh can be minimal or complete, and both are legitimate.** One with the small registry and no
+gitea builds and runs modules and cannot serve packages. That is a real configuration, not a broken
+one — the same way a partial host is real.
+
+**And the bootstrap still has no package registry.** The first build of the shared base happens
+before anything has installed one. That is the same pivot as everything else and it is **not solved
+here**: it is named, so the next person does not discover it.
+
+## How this is checked
+
+| Rule | Checked by |
+|---|---|
+| Genesis needs no database | The installer raises a mesh of one on a machine with nothing, and the artifact store it installs has no store of its own. |
+| Two providers are a choice, not a conflict | A mesh holding both is asked to resolve `artifact-store` and refuses, naming both, until one is assigned. |
+| Providers may share a machine | A node is assigned both gitea and a registry, and both run — only one of them answers `artifact-store`. |
+| The two stores are not interchangeable | A module depending on `package-registry` is not satisfied by `artifact-store`, and the refusal says why. |
+| A migration is verified before it is finished | The old provider cannot be unassigned while any machine's declaration still names it. |

02-DECISIONS/0076-the-sdk-is-a-published-package.md

@@ -0,0 +1,87 @@
+---
+topic: building it
+status: accepted
+date: 2026-09-16
+deciders: jochen
+reconstructed: false
+extends: 0075-two-stores-and-which-provides-what.md
+---
+
+# 76. The SDK is a published package, and the toolchain resolves it by version
+
+## Context
+
+[ADR 0014](0014-no-npm-workspace.md) decided a module consumes its dependencies — the mesh's own
+shared library included — from the private registry. [ADR 0075](0075-two-stores-and-which-provides-what.md)
+decided the private registry is a `package-registry` provision, and that gitea provides it. What
+neither settled, and what [issue 053](../04-ISSUES/053-the-sdk-is-pinned-twice-and-the-two-disagree/00-report.md)
+left open, is the one build where the rule cannot simply be obeyed: **the first one.**
+
+The TypeScript toolchain image is built *from* the SDK — it carries the SDK so that every module
+compiled inside it resolves the shared library without each build fetching it. So the thing that
+compiles TypeScript and the thing that contains the SDK were the same object, and that object
+cannot be what builds the SDK. Stated as a question — "how does the SDK reach the registry before
+the toolchain exists, when the toolchain is what builds it?" — it reads as a paradox.
+
+It is not one. The paradox exists only because the toolchain *bakes a git-cloned copy* of the SDK.
+The SDK itself is plain TypeScript: it needs `node` and `tsc` and nothing the mesh makes. A public
+base image can build it. The circularity is a property of the workaround, not of the SDK.
+
+## Decision
+
+**The SDK is an ordinary published package in the mesh's `package-registry`, consumed by version.**
+The git URL in the toolchain's manifest and the sibling-path lock beside it — the two halves of
+issue 053 — are both removed. A build resolves the SDK the way it resolves any dependency, with a
+lock that agrees with its manifest, so `npm ci` is the command and reproducibility is by
+construction rather than by the machine the build ran on.
+
+**The SDK is built with a public base image, not with the mesh's toolchain.** It is *not* one of
+the components the loop cannot build — the control plane, the registry, the builder, the catalogue
+([`12-a-module-repository`](../03-DESIGN/01-to-be/12-a-module-repository.md)), which arrive by
+carrying an init builder because they are the loop's own machinery. The SDK is machinery for
+nothing; it is an ordinary dependency the loop builds and publishes like any other. The only
+constraint is narrow: it cannot be compiled *in the mesh toolchain*, because that toolchain is built
+from it. So it is compiled on a public base image instead — which needs nothing the mesh makes — and
+published before the toolchain that consumes it. It is not carried, because building it does not
+wait on a mesh existing first.
+
+**The toolchain base stays, thinned.** mesh-tools remains the image bundles are compiled in and the
+one place the SDK is resolved — but it `npm ci`s the SDK by version from the registry instead of
+baking a copy cloned from a git URL. Bundles keep borrowing its resolved dependencies; what changes
+is that the version they borrow is named and honest. This was the shape chosen over dropping the
+shared base entirely and having every bundle resolve the SDK itself: one resolution point, one
+place to be right about the version.
+
+**Genesis orders the publish before the first compile.** The package-registry provider is a public
+image (gitea), so it comes up needing no toolchain; the SDK is published into it; only then is the
+toolchain built, so the first `npm ci` has a registry to read from. Nothing in that chain is
+circular, because the only thing that needed the toolchain — baking the SDK — is gone.
+
+## Consequences
+
+Each language's toolchain repeats the shape: its own SDK, built from that language's public base
+image, published to the same registry, resolved by version with that ecosystem's lockfile-honest
+install (`npm ci`, `cargo` against a vendored or registry source, `pip` against a pinned set). The
+warning in issue 053 — that whatever the TypeScript repository does the others will copy — is
+answered by making the copied thing the correct one.
+
+A change to the SDK is publish-then-consume, exactly as [ADR 0014](0014-no-npm-workspace.md) already
+priced it: publish the new SDK version, then bump the toolchain (and any module pinning it directly)
+to consume it. There is no shortcut that resolves an unpublished SDK, which is the property that was
+missing.
+
+mesh-tools is no longer an SDK carrier in the sense that mattered — it does not contain a copy
+whose provenance is a branch head somebody force-pushes. It contains a version.
+
+A mesh with no package-registry cannot build TypeScript. This is accepted and is not new: it is the
+same shape as a mesh that cannot reach a forge being unable to be raised
+([ADR 0071](0071-where-genesis-gets-its-source.md)). Installing brings the registry up first.
+
+## How this is checked
+
+| Rule | Checked by |
+|---|---|
+| The SDK a build compiles against is named, not cloned from a branch | The toolchain manifest pins `@novox/mesh-sdk` to a version, and the build runs `npm ci`, which refuses a lock that disagrees with the manifest. Issue 053's two checks become this one. |
+| The SDK builds without the mesh's own toolchain | The SDK's build recipe names a public base image. A recipe that named the mesh toolchain would reintroduce the cycle and is refused in review. |
+| The registry is up before the first compile | The genesis bed asserts the package-registry answers, and the SDK is published, before the base build runs. A base build that ran first would fail its `npm ci` with no registry, which is the positive control. |
+| A second language repeats the shape, not a new one | When a second SDK is added, its recipe is compared to this one: public base, publish by version, lockfile-honest install. |

02-DECISIONS/0077-the-controller-and-the-foundation.md

@@ -0,0 +1,61 @@
+---
+topic: the mesh
+status: accepted
+date: 2026-09-16
+deciders: jochen
+reconstructed: false
+extends: 0006-the-substrate-and-the-control-plane.md
+---
+
+# 77. The parts are named controller, foundation, node — not control plane, substrate, master
+
+## Context
+
+The words drifted. In conversation and in code the same thing was called *control plane*,
+*controller*, *master*, and *hub*; the store-and-broker pair was called *substrate* and
+*foundation*; a machine was a *node*, a *worker-node*, a *peer*, a *slave*. A mesh named
+differently by two people is a mesh they describe differently, and the drift was worst on the
+parts talked about most.
+
+Three of the terms carried wrong ideas. *Control plane* is borrowed from networking's
+control-plane/data-plane split and means nothing here. *Master/slave* and *hub/peer* imply a
+subordinate — but no node is: a node applies its own declaration and keeps running when the
+control-node dies, so it is as much its own machine as any other. *Substrate* is a biology
+metaphor that landed for no one.
+
+## Considered Options
+
+1. **Keep the inherited words.** Rejected: they are the source of the drift, and two of them
+   (control plane, substrate) are metaphors that teach the wrong shape to anyone reading them cold.
+2. **master / slave, or hub / peer, for the nodes.** Rejected: both name a hierarchy the mesh does
+   not have. The control-node owns no other node; lose it and the rest keep running what they were
+   last told.
+3. **controller / foundation / node + control-node.** Chosen.
+
+## Decision
+
+The component that decides what each node should be, holds the mesh's records, and tells nodes is
+the **controller** — the module `mesh-controller`, which claims the mesh-scoped `the-controller`
+seat. The store and broker raised at genesis are the **foundation**. Machines are **nodes**;
+there are 0..n of them, and exactly one — the one running the controller — is the **control-node**.
+
+Retired: *control plane*, *substrate*, *master/slave*, *hub/peer*, *worker-node*.
+[`00-META/glossary.md`](../00-META/glossary.md) is the authority, and a new name for an existing
+thing lands there in the change that introduces it in code.
+
+## Consequences
+
+`mesh-control` became `mesh-controller` across the module, container, image, binary, `cmd/` dir and
+the git repository; `substrate` became `foundation` in the embedded base bundles, the default
+template and the example lock; the seat `the-control-plane` became `the-controller`. The 03-DESIGN
+prose and 00-META follow the new words.
+
+What got harder: the records under `02-DECISIONS/` are immutable, so they keep the words they were
+written with — this record included, whose own title names what it retires. A term retired here
+still appears there, and the glossary is how to read it. The git repository on the forge was
+renamed `mesh-control` → `mesh-controller`.
+
+## References
+
+- [`00-META/glossary.md`](../00-META/glossary.md) — one name per thing, and the words retired.
+- The rename shipped across all six code repositories and hq (main).

02-DECISIONS/0078-the-store-and-broker-are-modules.md

@@ -0,0 +1,63 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-16
+deciders: jochen
+reconstructed: false
+extends: 0033-the-substrate-is-a-store-and-a-broker.md
+---
+
+# 78. The store and the broker are ordinary modules
+
+## Context
+
+[ADR 0033](0033-the-substrate-is-a-store-and-a-broker.md) settled that the foundation is a store
+and a broker, raised at genesis; [ADR 0006](0006-the-substrate-and-the-control-plane.md) settled
+that the controller cannot grant itself either, because it consumes them and is not running yet to
+ask. Both were raised as bundle resources — plumbing, with no record in the mesh's module graph.
+
+That left two costs, named in [issue 051](../04-ISSUES/051-the-mesh-cannot-update-what-it-depends-on/00-report.md).
+The foundation's own store and broker could not be upgraded — nothing owned them as modules. And a
+mesh that wanted a database or a queue for its modules installed the `postgres`/`lavinmq` modules,
+each of which raised a **second** server: a mesh ran two postgres and two brokers.
+
+## Considered Options
+
+1. **Leave them as bundle-only plumbing.** Rejected: they cannot be upgraded, and the second
+   server stays. The floor keeps a permanent specialty in it.
+2. **The control-plane pivot verbatim — raise a temporary one, install the module, retire the
+   temporary** ([ADR 0067](0067-genesis-is-a-pivot.md)). Rejected for a *stateful* server: it means
+   a handover with real downtime, tearing down the store the controller is mid-read of.
+3. **Adopt in place.** Chosen.
+
+## Decision
+
+The foundation's store and broker are **adopted in place** as the ordinary `postgres` and `lavinmq`
+modules. Genesis still raises them first (nothing else can — ADR 0006), then each module declares a
+container with the **same name, image and spec** the foundation raised, so the applier — which keys
+on the container name and compares a spec digest — reconciles it rather than raising a second. The
+credentials are the foundation's, made at genesis and carried in through `secret accept`, because
+the mesh cannot invent a credential that already made the databases. The servers bind mesh-wide so
+a consumer on any node can reach the one shared server.
+
+A mesh runs **one postgres and one lavinmq**, and each is upgradeable through a stated window: the
+store's is a connection-pool reconnect; the broker's is the harder case of recreating the bus the
+push travels over, so the mesh reconnects to the one that returns.
+
+## Consequences
+
+The twelve-module floor has no specialty left in it — the store and broker are moments in a
+module's life, not a separate kind of thing. Two follow-ups are tracked:
+[issue 054](../04-ISSUES/054-the-adopted-store-and-broker-are-open-before-the-filter/00-report.md)
+(the adopted servers bind `0.0.0.0` before the packet filter is installed) and
+[issue 055](../04-ISSUES/055-the-adopted-store-and-broker-may-be-reachable-on-one-node-only/00-report.md)
+(whether a consumer on another node reaches them over the overlay).
+
+What got harder: a foundation upgrade recreates the very server the controller reads from, or the
+bus the instruction to upgrade travels over — a window that a stateless module upgrade does not have.
+
+## References
+
+- [issue 051](../04-ISSUES/051-the-mesh-cannot-update-what-it-depends-on/00-report.md) — the gap this closes.
+- [`03-DESIGN/01-to-be/07-the-foundation.md`](../03-DESIGN/01-to-be/07-the-foundation.md) — the amended design.
+- Shipped across mesh-host, mesh-catalog and mesh-lab (main); proven 22/22 in the one-node lab.

02-DECISIONS/0079-the-foundation-seats-are-named-after-their-servers.md

@@ -0,0 +1,63 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-17
+deciders: jochen
+reconstructed: false
+extends: 0078-the-store-and-broker-are-modules.md
+---
+
+# 79. The foundation seats are named after their servers
+
+## Context
+
+[ADR 0078](0078-the-store-and-broker-are-modules.md) settled that the store and broker are the
+ordinary `postgres` and `lavinmq` modules, adopted in place on the control-node, and that a mesh
+runs **one postgres and one lavinmq**. But that singularity held only by convention: genesis
+assigns them to the control-node alone. [Issue 056](../04-ISSUES/056-an-adopted-module-assigned-to-a-second-node-raises-a-second-server/00-report.md)
+recorded the gap — a second `assign` to another node finds no container of that name there and
+raises a *second* server, holding none of the first's data, and nothing refuses it.
+
+The mesh already has the mechanism for "there is one of me": a mesh-scoped exclusive **seat**.
+[ADR 0077](0077-the-controller-and-the-foundation.md) gave the controller one, which it named
+`the-controller`. Nothing gave the store and broker theirs.
+
+## Decision
+
+Each foundation module claims a mesh-scoped seat named **after the server it guards**:
+`mesh-controller`, `mesh-store`, `mesh-broker`. So the `postgres` module claims `mesh-store` and
+the `lavinmq` module claims `mesh-broker`; the resolver refuses a second holder mesh-wide, the same
+way it keeps one hub and one controller. A second `assign` is now a refusal at resolution — *one
+per mesh* — not a silent second server.
+
+The controller's seat, which [ADR 0077](0077-the-controller-and-the-foundation.md) named
+`the-controller`, is renamed `mesh-controller` under this same rule, so all three foundation seats
+follow one convention: the seat is the server. (The module `mesh-controller` and its seat now share
+a name, which is the point — there is one of that server, and the seat says so.)
+
+This does not tie a foundation module to the control-node — a mesh-scoped seat forbids a *second*
+holder, not a wrong single one. Adoption still requires the container to already be running where
+the module lands, which genesis arranges; the seat closes the "two servers" gap, and the
+control-node convention remains what puts the one holder in the right place.
+
+## How this is checked
+
+- The resolver's `checkClaims` refuses two holders of a mesh-scoped seat
+  (`mesh-controller/internal/catalogue/resolve.go`).
+- `TestAFoundationModuleCannotBeRaisedOnASecondNode` asserts each foundation module's second
+  assignment is refused with *one per mesh*.
+- The `postgres`, `lavinmq` and `mesh-controller` manifests declare the seat.
+
+## Consequences
+
+"One store, one broker, one controller" is now a property the mesh enforces rather than a
+convention it hopes for. The silent operational edge that remains — a cross-node consumer
+provisioned only when the provider is pushed again — is separate, and tracked as
+[issue 057](../04-ISSUES/057-a-cross-node-consumer-is-provisioned-only-when-the-provider-is-pushed-again/00-report.md).
+
+## References
+
+- [issue 056](../04-ISSUES/056-an-adopted-module-assigned-to-a-second-node-raises-a-second-server/00-report.md) — the gap this closes.
+- [ADR 0077](0077-the-controller-and-the-foundation.md) — named the controller's seat, here renamed.
+- [ADR 0078](0078-the-store-and-broker-are-modules.md) — asserted one postgres and one lavinmq, here enforced.
+- [`00-META/glossary.md`](../00-META/glossary.md) — seat and claim.

02-DECISIONS/0080-the-development-cycle-is-checked.md

@@ -0,0 +1,55 @@
+---
+topic: how we work
+status: accepted
+date: 2026-09-17
+deciders: jochen
+reconstructed: false
+extends: 0019-how-this-repository-works.md
+---
+
+# 80. The development cycle is checked, not trusted
+
+## Context
+
+[ADR 0019](0019-how-this-repository-works.md) made this repository the source of truth, and the
+process overview drew the flow work must follow: an idea or a symptom, a decision, a to-be design,
+a build in a code repository, an as-is update on shipping. The playbooks describe every step, and
+frontmatter carries every status.
+
+But the flow itself was enforced by nothing. A design could appear citing no decision; a design
+could sit `in-progress` naming no code; an issue could be `fixed` by nobody knows what. Each is
+indistinguishable from correct work until somebody reads carefully — and the whole point of the
+playbooks is that nobody should have to hold this repository in their head. A session that starts
+cold (or an agent after a context clear) must be able to *find* the chain by following frontmatter
+pointers, which only works if the pointers are reliably there.
+
+## Decision
+
+The development cycle is enforced mechanically, to the extent frontmatter can carry it:
+
+- **No design without a decision** — every to-be design names at least one record in `decisions:`.
+- **No development without a design that says where** — an `in-progress` or `implemented` design
+  names its owning code in `code:`.
+- **No owner-less diagnosis, no fix-less fix** — an issue marked `located` or `fixed` names
+  `located-in:`; one marked `fixed` or `resolved` says `fixed-by:` (prose counts — "nothing, the
+  capability existed" is an answer).
+- **No silent graduation** — a `graduated` research overview says what it `became:`, and the
+  targets exist.
+
+[`00-META/checks/cycle.py`](../00-META/checks/cycle.py) refuses violations, beside `records.py`
+and `index.py`; all three run before any merge here. What frontmatter cannot see — that code work
+actually started from a handoff — remains held by playbooks 04 and 07: a feature branch exists
+because a design or an issue sent it, and a merge is a human checkpoint.
+
+## Consequences
+
+A `/clear` costs little: [`AGENTS.md`](../AGENTS.md) now carries the cycle and a where-to-look
+table, and the chain a fresh session needs is guaranteed present in frontmatter rather than
+reconstructed from memory. The checks are the floor, not the ceiling — they verify pointers exist,
+not that their content is true; reading remains the job.
+
+## References
+
+- [`00-META/process/00-overview.md`](../00-META/process/00-overview.md) — the flow, and its new
+  "The cycle is checked" section.
+- [ADR 0019](0019-how-this-repository-works.md) — the repository this disciplines.

02-DECISIONS/0081-a-decision-nothing-cites-is-not-yet-in-the-chain.md

@@ -0,0 +1,51 @@
+---
+topic: how we work
+status: accepted
+date: 2026-09-17
+deciders: jochen
+reconstructed: false
+extends: 0080-the-development-cycle-is-checked.md
+---
+
+# 81. A decision nothing cites is not yet in the chain
+
+## Context
+
+[ADR 0080](0080-the-development-cycle-is-checked.md) made the development cycle checked — but its
+checks covered designs, issues and research, not the decisions themselves. Measuring showed why
+that matters: 19 of 70 records were cited by nothing — no design doc's `decisions:`, no research
+`became:`, no issue, no other record. Among them sat load-bearing decisions (the credential flow,
+the module-runtime cluster), and the cost had already been paid once in practice: a stale premise
+about an orphaned decision survived in working memory precisely because no pointer led to the
+record that had settled it.
+
+## Decision
+
+Every **accepted** decision must be reachable from the cycle: cited by a design doc's frontmatter
+(`decisions:` — a governing citation, not a prose mention), a research overview, an issue report,
+a `00-META` document, or another record's `extends`/`supersedes` chain.
+[`cycle.py`](../00-META/checks/cycle.py) refuses orphans. Proposed records are exempt — a record
+under consideration has no home yet — and superseded records are reachable through their
+supersession chain by construction.
+
+The 19 orphans were given true homes in the same change: the module-runtime cluster
+(0044–0049, 0053–0055) into the connectivity, controller, protocol, writing and model-access
+designs; the build decisions (0072, 0076) into the building design; 0021 into the playbook that
+implements it; the process records were already reachable once `00-META` counted as a source.
+
+Taken together with 0080, the practice has a name the industry will recognise:
+**spec-driven development, with provenance** — a decision is the *why*, the design doc is the
+spec, `code:` names the implementation, and the lab beds are the conformance tests. What the
+common form leaves implicit, the cycle makes checked: the spec itself must trace to a decision,
+and the decision must be findable from the work it governs.
+
+## Consequences
+
+Following pointers now reaches every accepted decision, so a cleared session (or a person) can
+trust the frontmatter graph as the whole map. The check is reachability, not truth: a citation
+placed wrongly still lies, and reading remains the job.
+
+## References
+
+- [ADR 0080](0080-the-development-cycle-is-checked.md) — the cycle this completes.
+- [`00-META/process/00-overview.md`](../00-META/process/00-overview.md) — the flow.

02-DECISIONS/0082-the-registry-is-reached-by-name-and-trusted-by-the-overlay.md

@@ -0,0 +1,84 @@
+---
+topic: building it
+status: accepted
+date: 2026-09-17
+deciders: jochen
+reconstructed: false
+extends: 0072-two-graphs-and-the-build-chain.md
+---
+
+# 82. The registry is reached by name, and the overlay is its security
+
+## Context
+
+Delivery ends at a node pulling an image, and for every node but the one that built it, that step
+has never worked. Three facts conspired, each recorded separately:
+
+- Artifact references are written under `127.0.0.1:5000` — a deliberate parking (the catalogue
+  says so in the commit that reverted the mesh-reachable name: *"the runtime refused it: 'http:
+  server gave HTTP response to HTTPS client' … the binding expression returns then"*), so a joined
+  node is told to pull from its own loopback
+  ([issue 048](../04-ISSUES/048-nothing-makes-a-machine-trust-the-mesh-registry/00-report.md)).
+- A container runtime treats any non-loopback registry as HTTPS, and the mesh's registry serves
+  plain HTTP; nothing the mesh writes tells any runtime otherwise — the one place that file
+  existed was the lab's, which is why this never failed in a bed (issue 048).
+- Nothing gives a node an account for the registry
+  ([issue 042](../04-ISSUES/042-nothing-gives-a-node-an-account-for-a-registry/00-report.md)) —
+  and nothing has ever said whether one is required.
+
+The tempting fix is TLS from the mesh's own authority. The machinery even exists — the mesh CA
+issues leaves for `<node>.internal` names, and a `certificate:` manifest field delivers them. But
+the registry is reached **over the overlay**, and the overlay is WireGuard: every byte is already
+encrypted and already authenticated to a peer the mesh admitted. The mesh's own code has carried
+this position for weeks: *"the registry is reached over the mesh's own private network … a second
+layer inside it would be certificates to issue and rotate for no property the first does not
+have."* TLS inside the tunnel would also re-order genesis (a certificate needs the network, the
+registry precedes it) and put CA handling into three clients (the runtime, the archive fetcher,
+the builder) — cost with no new property.
+
+## Decision
+
+**The overlay is the registry's transport security, and the mesh writes the trust it means.**
+
+1. **References name the registry by its mesh name.** The builder publishes to
+   `<provider>.internal:<port>` (the binding's `at` and `port`) — the parked one-line change
+   lands. Genesis still publishes to loopback on the first machine, before any overlay exists;
+   references minted at genesis stay loopback and are valid where they matter — on that machine.
+   A rebuild re-pins to the mesh name.
+2. **Every node on the private network is told the mesh's registry speaks plain HTTP.** The
+   controller injects, into every such node's declaration, a merged `/etc/docker/daemon.json`
+   naming `<provider>.internal:<port>` under `insecure-registries`, and a service resource that
+   restarts the runtime when that file changes — the `/etc/hosts` pattern for the content, the
+   nftables pattern for the reload. No module author is involved; being on the network is what
+   grants the trust, because being on the network is what the trust *is*.
+3. **No accounts (issue 042), recorded as the position it always was.** Reading and pushing
+   require presence on the overlay and nothing else. The boundary is enforced, not assumed: the
+   registry's `listens` is `from: mesh`, the firewall derives from it, and the overlay admits only
+   peers holding mesh-issued keys. An operator's *external* registry credential is an ordinary
+   operator-supplied secret (`secret accept`), owned by whichever module names that registry.
+   Per-node accounts return as a decision, not a patch, if the boundary assumption ever changes —
+   the shape would be the existing provision flow.
+
+## How this is checked
+
+- The no-fake lab bed: a joined node pulls a mesh-built image by the registry's `.internal` name
+  over the overlay — an address outside every range the lab's own runtime configuration trusts,
+  so the mesh-written trust is what makes it work or nothing does.
+- The firewall half is generated from `listens` and visible in the node's ruleset (`from: mesh`).
+- The plaintext-inside-tunnel position holds exactly as long as the registry is unreachable off
+  the overlay; the `listens` stanza and the derived ruleset are the check.
+
+## Consequences
+
+A runtime restart when the trust first lands on a node — at joining, before workloads, where it
+is free. The registry's HTTP is exposed to whatever stands on the overlay, which is the stated
+boundary; a mesh that wants defence in depth inside its own tunnel reopens this record rather
+than bolting certificates on quietly. Issues 042 and 048 close on this record.
+
+## References
+
+- [issue 042](../04-ISSUES/042-nothing-gives-a-node-an-account-for-a-registry/00-report.md),
+  [issue 048](../04-ISSUES/048-nothing-makes-a-machine-trust-the-mesh-registry/00-report.md).
+- [ADR 0072](0072-two-graphs-and-the-build-chain.md) — the build chain this completes.
+- [ADR 0029](0029-a-network-is-a-shape-because-an-action-cannot-be-undone.md) — the overlay as a
+  boundary.

02-DECISIONS/0083-one-push-leaves-the-mesh-consistent.md

@@ -0,0 +1,59 @@
+---
+topic: the mesh
+status: accepted
+date: 2026-09-18
+deciders: jochen
+reconstructed: false
+extends: 0010-delivery.md
+---
+
+# 83. One push leaves the mesh consistent
+
+## Context
+
+A provision is minted while composing the *consumer's* node; the provider's grant list is a pure
+read of secrets already issued from it. So assigning a cross-node consumer and pushing its node
+produced a consumer that retried forever against a provider that had never heard of it, until the
+provider's node was pushed a second time — an action with no signal to take, documented nowhere,
+and invisible whenever consumer and provider share a machine
+([issue 057](../04-ISSUES/057-a-cross-node-consumer-is-provisioned-only-when-the-provider-is-pushed-again/00-report.md)).
+
+Two remedies were on the table: **cascade** — a push also delivers to the machines its compose
+changed — or **report** — a push says "now push the provider" and leaves the act to the operator.
+
+## Decision
+
+A push finishes what it starts: after composing and sending the named node, the controller flushes
+every *other* machine that is now behind — whose declaration differs from what it was last sent —
+by name, in the push's own output, converging over a bounded number of rounds (a flushed send may
+itself mint).
+
+Behind is measured against what a machine was last *sent*, not against a before/after snapshot of
+this push. The mint that makes a provider behind happens when the consumer is assigned or its
+account issued — before `push` runs at all — so by push time the provider already differs from
+what it holds, with no in-command delta to detect. The only durable signal is "what it should be"
+versus "what it last received", which is the same comparison `push --behind` already makes.
+
+A machine behind for an unrelated reason is flushed by this too, and that is correct rather than a
+cost: a named push that knew a machine was behind and left it so would be the very silence this
+decision removes. The narrower reading — flush only what this push provably changed — was
+rejected because it cannot see a mint that a prior command performed, which is precisely the 057
+case.
+
+Reporting alone was rejected because it converts a derived fact the controller already holds into
+an operator obligation, and an obligation enforced by nothing is issue 057 restated. The
+declaration is computed from the whole mesh; delivering a mesh that is knowingly inconsistent and
+merely saying so would make "push succeeded" mean less than it says.
+
+## Consequences
+
+- One push is sufficient for a cross-node consumer: the provider's grants arrive from the same
+  act that minted the provision. The undocumented rule "push the provider node too" ceases to
+  exist rather than becoming documentation.
+- A named push delivers to every machine that is behind, not only the one named — each named in
+  the output, never silent. `push --behind` remains the way to reconcile the mesh without naming
+  a node; a named push now carries the same guarantee for the machines its work touched and any
+  others already waiting.
+- How this is checked: the built-store-cross-node bed registers a cross-node consumer, pushes
+  only the consumer's node, and asserts the provider minted its vhost — the workaround push is
+  removed, so a regression fails the bed.

02-DECISIONS/0084-which-provider-serves-a-consumer.md

@@ -0,0 +1,108 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-20
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0027-a-provision-names-what-the-consumer-is-coupled-to.md
+---
+
+# 84. Which provider serves a consumer, when the mesh runs more than one
+
+## Context
+
+[ADR 0027](0027-a-provision-names-what-the-consumer-is-coupled-to.md) settled what a provision
+is *named* for — the thing the consumer's code is coupled to, so `postgres-database` and
+`mssql-database` are different provisions and a wrong match is refused at resolution. It said one
+thing more, in passing, and left it: *"Two providers of `postgres-database` — a container on this
+node and a managed instance elsewhere — are interchangeable and should both match."* Naming was
+decided; **which of several providers serves a given consumer was not.**
+
+The mesh assumes there is only one to choose. Provisions are mesh-scoped: the adopted store is a
+single `mesh-store` a consumer on any node reaches, and `scope: "mesh"` is written into the
+provision definitions. **That assumption is false on day one, and was always meant to be.**
+Node-specific services delivered to the mesh is the plan, not an edge case:
+
+- Both control-capable nodes already run their **own** general-purpose relational store (the same
+  engine, one instance each), serving that node's own applications.
+- Each runs its **own** SQL server, its **own** cache, its **own** object store. One node alone
+  runs six separate relational-store instances, each raised by the module that needed it.
+- The one provision that is currently single — the identity provider, one instance on one node —
+  already authenticates applications whose home is a **different** node.
+
+So several providers of one provision name genuinely coexist, and they are **not** interchangeable
+the way 0027's aside supposed. They differ by node, by the data they hold, and by locality. A
+consumer bound to the wrong one reads the wrong database, or takes a cross-node hop it did not
+need, or cannot be moved without silently rebinding. The model has no field in which to say which
+one. This is 0027's own fault — *a match that resolves and is wrong* — one level up: 0027 refused
+the wrong **dialect**; nothing refuses, or even asks about, the wrong **instance**.
+
+## Considered Options
+
+1. **Keep `scope: "mesh"` — one provider per provision, mesh-wide.** Rejected: it is false on day
+   one, and making it true would force every node's applications onto one node's server — the
+   exact opposite of node-specific services delivered to the mesh, and a single point of failure
+   the topology was built to avoid.
+2. **Resolve to any provider of the name (0027's "both match").** Rejected: when providers hold
+   different data and live on different nodes they are not interchangeable, and picking one
+   arbitrarily is a wrong-instance match — the confidently-wrong answer 0027 exists to prevent,
+   restated at the level of the instance rather than the dialect.
+3. **Always require the consumer to name the provider explicitly.** Rejected: needless ceremony in
+   the common case, where the consumer wants the provider on its own node; and a field every
+   manifest must carry is a field an author forgets, which then matches everything again — the
+   failure 0027 warned about for qualifiers.
+4. **A provision is node-scoped; the consumer selects the provider, defaulting to co-location.**
+   Adopted.
+
+## Decision
+
+**A provision is served by a provider identified by its node, and the consumer selects which one.**
+A provider is a (node, module) pair, not a mesh-wide singleton. A consumer's binding resolves to a
+specific provider, and the selection is part of the assignment
+([ADR 0046](0046-a-module-configuration-is-its-assignments-not-its-manifest.md)), not the
+manifest.
+
+**The default is co-location.** A consumer that names no provider is served by the provider of
+that provision on **its own node**. This is the common case and needs nothing said. A mesh with
+one provider of a kind is just the case where co-location and "the only one" coincide — expressed
+as *there happens to be one*, not as a scope.
+
+**A consumer coupled to a provider's data names it.** Where two consumers must share one database,
+or a consumer must reach a provider on another node, the assignment names that provider — because
+that coupling is exactly what may not be guessed, and naming it is what makes a later move safe.
+
+**A module need not consume a provider at all.** It may carry its **own** instance inside its own
+composition — on its own module network, publishing no host port, **not** declared as a provision —
+when a genuine engine fork or a pinned server version makes the shared provider unusable. Such an
+instance is invisible to resolution and can be bound by nothing else. The rule is *share by
+default; embed only when a fork or a version forces it* — most of the per-module stores that exist
+today are vanilla engines on stale pins that a consolidation onto the node's provider would absorb.
+
+## Consequences
+
+The mesh can carry its real topology **deliberately** rather than by the accident of which
+provider happened to be the single one. A consumer's data-coupling becomes a stated fact, which is
+what lets a provider be moved without a consumer silently following the wrong one — provided a
+provider keeps its identity across a relocation, which is a follow-up this record opens rather than
+closes. Rotation ([13](../03-DESIGN/01-to-be/13-credentials-and-their-rotation.md)) addresses a
+specific provider's holders, not "the provision's".
+
+What got harder: an assignment now may carry a provider selection, and a wrong one is a new way to
+misconfigure. It is mitigated the way 0027 mitigated its own: the co-location default removes the
+choice in the common case, and genuine ambiguity — several providers, none named, none co-located —
+is refused with the candidates named, never resolved by picking.
+
+## References
+
+- [ADR 0027](0027-a-provision-names-what-the-consumer-is-coupled-to.md) — the naming this extends;
+  its "both match" aside is the gap closed here.
+- [ADR 0031](0031-the-control-plane-authenticates-nobody.md) — identity is exactly such a
+  provider, and already serves consumers on another node.
+- [ADR 0046](0046-a-module-configuration-is-its-assignments-not-its-manifest.md) — where the
+  selection lives.
+- [ADR 0078](0078-the-store-and-broker-are-modules.md) — the adopted store, whose mesh-wide
+  binding is the single-provider assumption this record replaces.
+- [issue 067](../04-ISSUES/067-a-provision-cannot-name-which-provider-serves-it/00-report.md) — the
+  gap, and the day-one evidence.
+- [`03-DESIGN/01-to-be/23-choosing-a-provider.md`](../03-DESIGN/01-to-be/23-choosing-a-provider.md)
+  — the design.

02-DECISIONS/0085-a-secret-is-a-provision.md

@@ -0,0 +1,153 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-20
+amended: 2026-09-20
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0031-the-control-plane-authenticates-nobody.md
+---
+
+# 85. A secret is a provision, and the vault is the module that provides it
+
+## Context
+
+[ADR 0031](0031-the-control-plane-authenticates-nobody.md) decided that identity *runs on the
+mesh, not of it* — a module other modules require, rather than a privileged part of the
+controller. [ADR 0078](0078-the-store-and-broker-are-modules.md) did the same for the store and
+the broker: the twelve-module floor has no specialty left in it. **Secrets are the exception that
+survived.** No module owns a secret.
+
+Secret handling is smeared across three built-in parts of the runtime:
+
+- the **controller mints** one credential per consumer↔provider pair and seals it to both node
+  keys ([ADR 0048](0048-a-provider-creates-the-credential-the-mesh-minted.md));
+- the **mesh generates local secrets** — *"generated secrets are the mesh's, never authored"*
+  (as-is, `06-configuration-and-secrets.md`) — for a value a single module needs for its own use;
+- the **synchroniser injects** both into a node's generated files.
+
+Nothing is the owner of "a secret" the way the store module is the owner of "a database", and the
+cost is recorded rather than hypothetical. The as-is design names the weakness in its own words:
+*"Rotation is not a mesh operation… there is no mechanism that rotates one and informs everything
+holding it."* A `rotate` command has since been built and proven, but it reaches **only** the
+provisioned pairs; a secret a module generates for its own fully-local use — the password of a
+version-pinned embedded store ([ADR 0084](0084-which-provider-serves-a-consumer.md)), an internal
+token — is minted by the mesh and then has no operation that can remake it. Three species of
+secret, and only the first has an owner:
+
+| species | minted by | rotates? |
+|---|---|---|
+| a provisioned credential (a database login) | controller, sealed to nodes (0048) | yes — `rotate`, per pair |
+| a module's own local secret | the mesh, as a generated value | **no owner, no rotation** |
+| an operator-delivered secret (an external key) | a person, sealed in (`secret accept`) | no rotation, no audit |
+
+## Considered Options
+
+1. **Leave it a property of the controller.** Rejected: it is the smear above — no owner, local
+   secrets that cannot be rotated, operator secrets that cannot be audited — and it is exactly the
+   specialty 0078 removed for the store and broker, kept here for no reason anyone recorded.
+2. **A dedicated vault built into the foundation, not a module.** Rejected: it reintroduces a
+   privileged built-in, the thing 0031 and 0078 went out of their way to remove, and a mesh that
+   wants none would still carry it.
+3. **Fold all minting, the controller's provisioning credentials included, into the vault.**
+   Rejected: the controller must mint in order to **deliver** any provision — the vault's own
+   credential among them — so making the vault mint the credential of its own delivery is the
+   store/broker chicken-and-egg for no gain. The provisioned-pair credential already has an owner
+   and a rotation ([13](../03-DESIGN/01-to-be/13-credentials-and-their-rotation.md)); this record
+   does not disturb it.
+4. **A secret is a provision; the vault is an ordinary module that provides it.** Adopted.
+
+## Decision
+
+**Secret-holding is a module, parallel to identity.** A module that needs a secret **for its own
+use** — a local service's password, an internal token, an external key it was handed — requires a
+`secret` provision from a vault provider, exactly as it requires a database from the store. The
+vault generates the value (or holds one it was given), and because the credential belongs to the
+consumer↔vault pair it **rotates, backs up and is audited through the same per-pair machinery**
+[13](../03-DESIGN/01-to-be/13-credentials-and-their-rotation.md) already defines. That is what
+gives the second and third species the rotation and audit they lack. A module's own secret stops
+being a generated value that nothing owns and becomes an ordinary provision with a provider.
+
+**The controller's minting of provisioning-pair credentials is unchanged** ([ADR 0048](0048-a-provider-creates-the-credential-the-mesh-minted.md)):
+it is how every provision, the vault's own included, is delivered. The vault does not mint the
+mesh's delivery credentials; it provides secrets to modules, and is itself provisioned the ordinary
+way.
+
+**The vault is node-scoped like every provider** ([ADR 0084](0084-which-provider-serves-a-consumer.md)):
+each node its own, selected the same way, so a module's own secret is held by the vault on the
+module's node. **A mesh that wants no vault runs none** — a module requiring no secret needs
+nothing, which is the same test 0031 applied to identity.
+
+## Consequences
+
+The gap that opened this — a generated local secret with no rotation — **closes without new
+machinery**: rotating such a secret is the vault's provisioner remaking a pair credential, the
+operation 13 already specifies. Backup, audit and break-glass gain an owner — the vault module —
+and become things a design specifies rather than absences. The as-is sentence *"rotation is not a
+mesh operation"* is already false for provisioned pairs and, once the vault ships, for local
+secrets too; the as-is document is updated when it does, not before.
+
+What got harder: a break-glass path — recovering a secret when the sealed delivery path is
+unavailable — must not reintroduce a key that one place holds, which is the property
+[ADR 0048](0048-a-provider-creates-the-credential-the-mesh-minted.md) was built to preserve; how
+the vault offers recovery without it is left to the design as an open question. And a module that
+today bakes a password into its own composition must instead require it from the vault — a
+migration taken module by module, not a flag day.
+
+## Amendment — 2026-09-20, before anything shipped
+
+Recorded on the record itself rather than as a supersession, by its decider, on the day it was
+accepted and before any code was merged against the sentences that change. The original text above
+is left as written; this section says what it got wrong and what stands instead.
+
+**What it got wrong.** The decision treated the vault as a provider like the store — optional, and
+one per node — and left the mesh's own root secrets outside it: the store's superuser, the broker's
+administrator, the controller's contexts, sealed to a node key and nothing else. Those are the
+secrets with no rotation and no recovery, and they are the ones a vault exists for. At genesis they
+are not even secret: the foundation raises its store and broker with fixed, well-known credentials
+and carries those into the mesh. Leaving that floor in place gave module secrets an owner and the
+root secrets none.
+
+**What stands instead.**
+
+- **The vault is a foundation module.** It is installed at genesis as part of the foundation
+  ([ADR 0078](0078-the-store-and-broker-are-modules.md) is the precedent: a foundation piece is
+  still an ordinary module), not assigned later by a mesh that happens to want one. *"A mesh that
+  wants no vault runs none"* is withdrawn. A mesh has root secrets, so a mesh has a vault.
+- **One per mesh, on the control-node.** *"The vault is node-scoped like every provider"* is
+  withdrawn. Node scoping ([ADR 0084](0084-which-provider-serves-a-consumer.md)) exists because a
+  store holds data a consumer is coupled to; the vault holds nothing a consumer is coupled to, and a
+  second one would be a second place to lose. Consumers on other nodes reach it as they reach
+  identity.
+- **The vault holds the mesh's root secrets under an operator-held key.** The controller mints and
+  delivers exactly as before; in addition, every secret a module holds for itself is sealed a second
+  time, to an **operator sealing key** whose private half never enters the mesh. The vault keeps
+  those operator-sealed copies on its own disk, outside the store, and can hand them out — they are
+  ciphertext to everything but the operator. This is the break-glass path the original text left
+  open, and it does **not** reintroduce a key one place holds: the mesh holds blobs it cannot open,
+  and the operator holds a key with nothing to open until given a blob. Recovery needs both.
+- **Genesis mints real root secrets and seals them to the operator key first**, so the fixed
+  credentials the foundation is raised with are replaced before the mesh is handed over.
+
+**Unchanged.** A module's own secret is a `secret` provision the controller mints and the vault
+records; the provisioned-pair path of [ADR 0048](0048-a-provider-creates-the-credential-the-mesh-minted.md)
+is untouched; the vault stores no plaintext, ever.
+
+**What it costs.** An operator key is a thing a person must keep, and a mesh whose operator key is
+lost has root secrets that can be rotated but not recovered — the same standing as today, stated.
+Sealing every own secret twice is a column and a call. Genesis grows a step. A module's
+vault-provided secret (a pair credential) is not yet sealed to the operator key; that is the next
+increment, not this one.
+
+## References
+
+- [ADR 0031](0031-the-control-plane-authenticates-nobody.md) — identity is a module; this is the
+  same move for secrets.
+- [ADR 0048](0048-a-provider-creates-the-credential-the-mesh-minted.md) — the provisioning-credential
+  path, left unchanged.
+- [ADR 0078](0078-the-store-and-broker-are-modules.md) — the de-specialisation this completes.
+- [ADR 0084](0084-which-provider-serves-a-consumer.md) — the node-scoping the vault obeys.
+- [issue 068](../04-ISSUES/068-secrets-have-no-owning-module/00-report.md) — the gap.
+- [`03-DESIGN/01-to-be/24-the-secrets-vault.md`](../03-DESIGN/01-to-be/24-the-secrets-vault.md) —
+  the design. The source mesh's `secret_locate` / `secret_backup` / `secret_verify` /
+  `secret_breakglass` subsystem is the prior art it draws on.

02-DECISIONS/0086-a-secret-reaches-a-process-as-a-file.md

@@ -0,0 +1,75 @@
+---
+topic: building it
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0085-a-secret-is-a-provision.md
+---
+
+# 86. A secret reaches a process as a file, and an exception is declared
+
+## Context
+
+The mesh seals a secret to the machine that uses it and discards the plaintext; the host unseals it
+into a file at 0600 ([ADR 0048](0048-a-provider-creates-the-credential-the-mesh-minted.md)). Then
+the module hands it to its container through an env-file, and the runtime puts it where anything
+on the machine that can talk to the runtime can read it: `docker inspect` prints it, and
+`/proc/<pid>/environ` holds it for the life of the process
+([issue 041](../04-ISSUES/041-a-sealed-credential-ends-up-in-the-process-environment/00-report.md)).
+
+Two facts made this a decision rather than a fix. **It is a supported shape**: placing
+`${secret:…}` in a file a container reads as its environment is what playbook 06 shows, and 36
+containers in 25 catalogue modules do it, the controller's own among them. And **nothing says which
+secrets are exposed**: a reader cannot tell from a manifest whether a credential is protected from
+`inspect` or not, because the manifest looks the same either way. The vault
+([ADR 0085](0085-a-secret-is-a-provision.md)) rests on the seal this undoes.
+
+## Considered Options
+
+1. **Leave it: the environment is where configuration goes.** Rejected — it gives the sealed value
+   back to a routine operation, and the care taken to seal it is then theatre.
+2. **Refuse every secret in an environment.** Rejected — some software reads its configuration
+   from the environment and nothing else, and a rule the catalogue cannot obey is a rule that gets
+   switched off.
+3. **A secret reaches a process as a file; an environment exception is declared, with a reason,
+   and refused otherwise.** Adopted.
+
+## Decision
+
+**A secret reaches a process as a file.** A module mounts the file the host wrote and points the
+program at it; the mesh's own programs accept a `_FILE` twin for every variable that carries a
+credential, the way the controller's store connections already did.
+
+**A container that reads a secret from its environment says so.** The manifest key
+`secrets-in-environment` on the container carries the reason. It is catalogue-level — the host
+never sees it — and it exists so a reader can tell from the manifest which secrets are exposed
+that way and why.
+
+**Everything else is refused.** A `${secret:…}` placeholder inside a container's `env` is never
+filled and is refused outright. A file carrying a secret that a container names in `env-file` is
+refused unless the container declares the exception.
+
+## Consequences
+
+The property *a sealed credential is readable only where it is used* becomes a manifest-level
+fact: true where no exception is declared, stated where one is. The controller reads all six of
+its credentials from files; the 35 other containers are marked with their reason and convert one
+by one where their software accepts a path, which is per-module work under
+[issue 041](../04-ISSUES/041-a-sealed-credential-ends-up-in-the-process-environment/00-report.md).
+
+What got harder: a module author meets one more refusal, and the reason they write is only as
+honest as they are. What is not decided: the reach of the exposure on a node — which identities can
+talk to the runtime — which decides whether a declared exception is a hardening item or something
+sharper.
+
+## How it is checked
+
+The catalogue engine refuses at composition, before anything reaches a machine, and its tests
+refuse both shapes and prove the reason is stripped from what is sent.
+
+## References
+
+- [issue 041](../04-ISSUES/041-a-sealed-credential-ends-up-in-the-process-environment/00-report.md)
+- [ADR 0048](0048-a-provider-creates-the-credential-the-mesh-minted.md), [ADR 0085](0085-a-secret-is-a-provision.md)
+- [`03-DESIGN/01-to-be/13-credentials-and-their-rotation.md`](../03-DESIGN/01-to-be/13-credentials-and-their-rotation.md)

02-DECISIONS/0087-a-seeded-file-is-created-once.md

@@ -0,0 +1,63 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0010-delivery.md
+---
+
+# 87. A seeded file is created once, and what grows in it is not the mesh's
+
+## Context
+
+A declaration is complete for what the host owns, and the host reconciles what is declared
+([ADR 0010](0010-delivery.md)): a file with this content, held to it. That is the only thing a
+manifest could say about a file, and it is the wrong thing for a file a module needs to **exist
+before first start** and something else then legitimately writes into — an access list a
+provisioner appends consumers to and the program persists back, a bootstrap configuration a
+program rewrites. Every reconcile restored the seed behind the running program, erased what had
+grown in it, and reported success
+([issue 035](../04-ISSUES/035-reconciling-a-seed-file-wipes-what-grew-in-it/00-report.md)).
+
+A run-once step ([ADR 0052](0052-a-step-that-runs-once-before-a-container.md)) can write a seed
+only if absent, and that closed the instance for a broker whose seed is a program's job. It left
+the general case: a plain file the mesh writes and never overwrites.
+
+## Considered Options
+
+1. **Two owners never share a file: the provisioner owns it, and first-start ordering is solved
+   another way.** Rejected as the only answer — some software refuses to start without the file,
+   and a module that must ship a program merely to write an empty file has been made to write a
+   program to say one word.
+2. **A create-once semantic on a file.** Adopted.
+
+## Decision
+
+A file resource may say `create-once`. The host writes it when it is absent and, when it is
+present, leaves it entirely alone — content, mode and owner — and reports it as **kept**, not
+corrected. What is in the file then is somebody else's work the mesh asked for. The mesh removes
+nothing it did not create ([ADR 0030](0030-data-outlives-the-mesh-that-declared-it.md)); it now
+also does not overwrite what it created once and handed over.
+
+On the security question ADR 0010 asks of every new resource behaviour: this **narrows** what a
+declaration can do to a machine. A create-once file gives a compromised control plane one fewer
+way to change a machine repeatedly — it can seed, once, and never again.
+
+## Consequences
+
+A module says which of its files are seeds, and the difference is visible in the manifest rather
+than in whether the file happened to be revisited. A later change to a seed's declared content
+does not reach a machine that already has the file; that is the meaning of a seed, and a module
+that needs the new content ships it as a run-once step that migrates the existing file.
+
+## How it is checked
+
+The host's apply tests: a seed is created, grown into by hand, reconciled, and the growth survives
+with the outcome `kept`. The vault bed declares one on a real node, grows it, pushes again, and
+reads it back.
+
+## References
+
+- [issue 035](../04-ISSUES/035-reconciling-a-seed-file-wipes-what-grew-in-it/00-report.md)
+- [ADR 0010](0010-delivery.md), [ADR 0030](0030-data-outlives-the-mesh-that-declared-it.md), [ADR 0052](0052-a-step-that-runs-once-before-a-container.md)

02-DECISIONS/0088-the-foundation-filters-before-anything-listens.md

@@ -0,0 +1,60 @@
+---
+topic: the mesh
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0078-the-store-and-broker-are-modules.md
+---
+
+# 88. The foundation filters before anything listens
+
+## Context
+
+Adopting the store and broker as ordinary modules
+([ADR 0078](0078-the-store-and-broker-are-modules.md)) needs them reachable by consumers across
+the mesh, so genesis raises them bound to every interface. The packet filter that decides who may
+reach them is a module too, installed a dozen steps later. Between the two, a control-node facing
+the network has its store and its bus open to anyone who can reach the machine, for the length
+of the install ([issue 054](../04-ISSUES/054-the-adopted-store-and-broker-are-open-before-the-filter/00-report.md)).
+The design's rule — what a port is reachable from is decided by the filter — is enforced by
+nothing for that window.
+
+## Considered Options
+
+1. **Accept the window**: the machine is mid-bootstrap and the exposure matches what the
+   pre-adoption modules had in steady state. Rejected — the whole point of deriving the filter
+   was to stop accepting that.
+2. **Bind narrowly at genesis and widen once the filter exists.** Rejected — a bind change is a
+   recreate of the mesh's store during install, and adoption in place needs the same spec.
+3. **The foundation carries a filter of its own, applied before the store.** Adopted.
+
+## Decision
+
+The foundation bundle installs the packet filter and loads a base ruleset **before the store and
+broker are raised**: drop by default; keep loopback, replies, ping and ssh; keep the mesh's own
+ports a node must reach before it is on the private network — the bus it enrols over and the
+registry it pulls from; and let the container runtime's own networks through the forward chain so
+containers keep working. It is written into the **same table** the filter module later derives, so
+that module replaces it wholesale the moment it can compute one from what the mesh knows, and
+nothing of the base survives to contradict it.
+
+## Consequences
+
+From its first resource a machine being made into a mesh refuses what it will refuse when
+finished; the window closes. What got harder: the base ruleset is static and names two ports the
+mesh's derived one also names — a change to which ports the foundation needs is now made in two
+places, and the bundle's own test says which.
+
+## How it is checked
+
+The installer's bundle test asserts the filter and its load precede the store and broker and that
+the rules name ssh, the bus and the registry and not the store's or broker's client ports. The
+genesis bed probes the machine from outside throughout the install: the store's port is never
+reachable, while the bus becomes reachable.
+
+## References
+
+- [issue 054](../04-ISSUES/054-the-adopted-store-and-broker-are-open-before-the-filter/00-report.md), issue 047
+- [ADR 0078](0078-the-store-and-broker-are-modules.md)
+- [`03-DESIGN/01-to-be/07-the-foundation.md`](../03-DESIGN/01-to-be/07-the-foundation.md)

02-DECISIONS/0089-a-bed-reads-the-catalogue-it-proves.md

@@ -0,0 +1,69 @@
+---
+topic: checking it
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0016-the-lab.md
+---
+
+# 89. A bed reads the catalogue it proves
+
+## Context
+
+A lab bed installs a catalogue module and asserts what the mesh does with it; "proven in the
+lab" is the standard a module must meet before it ships. The beds built the manifests they
+install inline — a literal copied from the catalogue when each bed was written, and never since.
+Six modules were converted to file-delivered secrets ([ADR 0086](0086-a-secret-reaches-a-process-as-a-file.md))
+and not one bed ran the converted shape; each ran its copy, and the copies still delivered
+secrets the way the catalogue engine now refuses
+([issue 073](../04-ISSUES/073-beds-carry-copies-of-catalogue-manifests/00-report.md)). A run's
+receipt named the commits of the host, the controller and the lab, and said nothing about the
+catalogue, so a catalogue change and a proven catalogue change were indistinguishable.
+
+The end-to-end design already has the rule this breaks — *the run rebuilds what it tests; an
+artifact rebuilt from memory is one rebuilt sometimes* — for binaries and images. A manifest is
+an artifact too.
+
+## Considered Options
+
+1. **Keep the copies and check them against the catalogue** — a test that diffs each literal
+   against the module's manifest, ignoring what the lab must rewrite. Rejected: it keeps two
+   sources of truth and adds a third thing that can drift, the list of what to ignore.
+2. **Read the catalogue, rewriting only what the lab must.** Adopted.
+
+## Decision
+
+A bed that installs a catalogue module reads that module's manifest from the catalogue checkout
+the run was pointed at. It may rewrite what the lab must and nothing else: a build artifact
+becomes the image the machine holds, an image is pinned to what the machine holds, a host port
+is remapped where one machine carries colliding modules, and an address may point at a stand-in
+the bed raises in place of an upstream. Everything else is the catalogue's, verbatim.
+
+A bed that needs less than the catalogue declares — no upstream server, a secret in the
+environment, a requirement edge removed — is not testing that module. It is a mesh test, and it
+carries a name of its own (see [issue 074](../04-ISSUES/074-a-mesh-test-wears-a-catalogue-modules-name/00-report.md)).
+
+The run's receipt names the catalogue's commit alongside the other repositories', so a receipt
+taken before a manifest changed says so.
+
+## Consequences
+
+A catalogue change is proven by the beds that install the module, or it is not proven, and the
+receipt says which. What got harder: a bed can no longer trim a module to the shape it finds
+convenient; it meets the module's declared requirements or gives its fixture another name.
+The beds that still carry a copy are declared, each with its reason, and the declared list
+only shrinks.
+
+## How it is checked
+
+A unit test in the lab refuses an inline manifest literal that names a catalogue module unless
+the bed is declared, with its reason, in the test's own list; a declaration for a bed that no
+longer carries the copy is refused too, so the list cannot outlive the debt. The receipt test
+asserts the catalogue is claimed whenever the run is pointed at one.
+
+## References
+
+- [issue 073](../04-ISSUES/073-beds-carry-copies-of-catalogue-manifests/00-report.md), [issue 074](../04-ISSUES/074-a-mesh-test-wears-a-catalogue-modules-name/00-report.md)
+- [ADR 0016](0016-the-lab.md), [ADR 0086](0086-a-secret-reaches-a-process-as-a-file.md)
+- [`03-DESIGN/01-to-be/01-end-to-end-testing.md`](../03-DESIGN/01-to-be/01-end-to-end-testing.md)

02-DECISIONS/0090-a-failure-that-repeats-is-said-to-be-stuck.md

@@ -0,0 +1,65 @@
+---
+topic: the mesh
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0010-delivery.md
+---
+
+# 90. A failure that repeats is said to be stuck
+
+## Context
+
+Delivery is a comparison, not a one-shot ([ADR 0010](0010-delivery.md)): a node re-applies the
+declaration it holds on a steady interval and reports each time. That is right for a failure that
+goes away by itself — the overlay not up yet, a registry briefly unreachable — and it makes a
+failure that will never go away look exactly the same. A resource nothing can ever apply is
+attempted, fails, is reported, and is attempted again every few minutes, indefinitely; the mesh
+keeps one report per machine, replaced, so each attempt arrives as "failed" at a fresh time.
+Nothing distinguished "failed once, will succeed when its dependency arrives" from "failed
+identically for ever", and nothing escalated the second
+([issue 065](../04-ISSUES/065-a-permanently-failing-resource-is-retried-for-ever-with-no-escalation/00-report.md)).
+
+## Considered Options
+
+1. **The host gives up** after some number of attempts. Rejected: the host does not know whether
+   a failure is permanent — that a registry has not answered three times is not evidence it never
+   will — and a host that stops trying is a node that must be pushed to again by hand.
+2. **A duration** since the failure was first seen. Rejected as the signal: a laptop shut for a
+   week has had one attempt, and a week is not evidence of anything.
+3. **The controller counts identical reports**, and says when there are enough of them. Adopted.
+
+## Decision
+
+The mesh keeps, beside each machine's last report, when the current failure was first reported
+and how many reports in a row have said it — the same outcome, the same refusal, the same failed
+resources by id. Not by the host's words: an error carrying a duration or a counter would read as
+new on every report, and the resource looping on it is exactly what this is for. A report that
+says something different starts the count again; a clean apply clears it. **Three identical reports in a row make a machine stuck**: `status` says
+so beside the failure, with the count and the time it began, and the machine-readable status
+carries the same three facts. The host keeps retrying; being stuck is a statement about the
+mesh's knowledge, not an instruction to the machine.
+
+The controller counts rather than the host, because only it sees every node: one stuck machine
+and a mesh-wide fault are different situations, and the host cannot tell them apart.
+
+## Consequences
+
+A resource that will never apply is visible from `status` after three reconcile intervals, to
+anyone who looks, without being asked for. What got harder: nothing on the machine changes — a
+gating failure that stops what follows still stops it, and this only makes the wait visible.
+Whether a stuck machine should also be raised as an event, and what a gating failure should do,
+stay open in the issue's own questions.
+
+## How it is checked
+
+An inventory test records the same failure three times and asserts the count and the unchanged
+start; the same resource failing in other words, and asserts the count went on; a different
+failure, and asserts it restarted; a clean apply, and asserts both cleared. The status command's own test asserts a stuck machine is said to be one.
+
+## References
+
+- [issue 065](../04-ISSUES/065-a-permanently-failing-resource-is-retried-for-ever-with-no-escalation/00-report.md)
+- [ADR 0010](0010-delivery.md)
+- [`03-DESIGN/01-to-be/10-delivery.md`](../03-DESIGN/01-to-be/10-delivery.md)

02-DECISIONS/0091-a-mount-is-declared-three-ways.md

@@ -0,0 +1,68 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0051-shared-data-is-the-operators.md
+---
+
+# 91. A mount is declared, and there are three things it can be
+
+## Context
+
+A container's bind mount whose source does not exist is created by the container runtime, as
+root, with whatever mode it picks. So `owner` and `mode` — which exist so a module can say who
+its data belongs to — never reach the directories that hold data, and the rule that keeps a
+directory when a module goes away ([ADR 0030](0030-data-outlives-the-mesh-that-declared-it.md))
+does not cover them, because the mesh has never heard of them
+([issue 026](../04-ISSUES/026-the-data-directories-are-not-declared/00-report.md)). Fourteen
+such mounts were declared by hand; a check that every mount is declared was then written and
+withdrawn, because it refused the builder: the builder mounts the container runtime's socket,
+which is not its data, already exists, and belongs to the machine. Declaring it as the module's
+own directory would be a lie the host would act on.
+
+## Considered Options
+
+1. **Declare the socket as a directory anyway.** Rejected: the host would create, own and
+   protect a path that is the machine's.
+2. **A new manifest field** naming machine paths a module may mount. Rejected: the manifest
+   already says the module needs the container runtime, and a second field would say the same
+   thing in paths.
+3. **Three declarations, one for each kind of path a mount can be.** Adopted.
+
+## Decision
+
+A container may not mount a path the module never declared, and a path is declared in one of
+three ways, which are the three things a path can be:
+
+- **the module's own** — a directory or file resource, or where a secret, a grant or a
+  contribution lands. Created and owned by the mesh for this module, kept when the module goes;
+- **the operator's** — an `accesses` entry ([ADR 0051](0051-shared-data-is-the-operators.md)):
+  pre-existing, shared, granted for use, never owned;
+- **the machine's** — a facility a declared capability grants. `container-runtime` grants its
+  socket. The path exists, the machine owns it, and the capability is the declaration.
+
+A mount under a declared directory is declared. The check runs where the manifest is parsed,
+and names the path and the three remedies.
+
+## Consequences
+
+Every directory that holds a module's data is one the mesh created with the module's owner and
+mode, and one ADR 0030 protects. What got harder: a manifest borrowed from a compose file no
+longer passes on the strength of its volume lines; each must say what kind of path it mounts.
+The table of what a capability grants is small and in the catalogue's parser; a new capability
+that grants a path adds a row.
+
+## How it is checked
+
+Manifest tests refuse an undeclared mount, accept one under a declared directory, accept one
+the module accesses, and accept the runtime's socket with the capability and refuse it without.
+A test parses every manifest in the catalogue beside the checkout and fails on any that breaks
+the rule, so the catalogue cannot drift back.
+
+## References
+
+- [issue 026](../04-ISSUES/026-the-data-directories-are-not-declared/00-report.md)
+- [ADR 0030](0030-data-outlives-the-mesh-that-declared-it.md), [ADR 0051](0051-shared-data-is-the-operators.md)
+- [`03-DESIGN/01-to-be/18-building-a-module.md`](../03-DESIGN/01-to-be/18-building-a-module.md)

02-DECISIONS/0092-an-operator-delivers-a-pair-credential.md

@@ -0,0 +1,61 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0085-a-secret-is-a-provision.md
+---
+
+# 92. An operator delivers a pair credential, and the mesh never replaces it
+
+## Context
+
+[ADR 0085](0085-a-secret-is-a-provision.md) names three species of secret and gives the vault
+two of them: a module's own secret, which the mesh mints, and an operator-delivered secret — a
+credential for something outside the mesh, which only a person can supply. Under 0085 a secret
+from the vault is a pair credential between the consumer and the vault. The controller's one
+command that takes a value from a person wrote only a module's own secret, sealed to one node.
+Nothing could put a value into a pair, so the third species had no entry
+([issue 070](../04-ISSUES/070-an-operator-cannot-deliver-a-pair-credential/00-report.md)), and
+an operator's credential could be held only as an own secret — un-audited, un-rotatable, the
+gap 0085 opened to close.
+
+## Considered Options
+
+1. **A new verb on the pair.** Rejected: `secret accept` already means "a value a person
+   supplied, sealed on the way in, plaintext discarded"; a second verb would mean the same.
+2. **`secret accept` grows a provider end.** Adopted.
+
+## Decision
+
+`secret accept <node> <module> <name> --provider <node>` seals the supplied value to the
+consumer's node, to the provider's node, and to the operator's key when the mesh has one, and
+records the pair as `accepted`. Every pair credential now says where it came from: `made` or
+`accepted`.
+
+An accepted pair is never replaced by a made one. When a sealing key at either end changes, the
+mesh cannot re-seal a value it does not hold, so the read is refused and names the remedy —
+accept it again. `rotate` refuses an accepted pair for the same reason: the mesh cannot make its
+replacement, and deleting it would have the next read mint one, delivered and reported as
+applied while failing to authenticate somewhere else entirely. Rotating an accepted credential
+is accepting a new value.
+
+## Consequences
+
+A credential for something outside the mesh lives in the vault's ledger with the others, sealed
+to both ends and recoverable by the operator. What got harder: a mesh whose node keys change
+cannot heal an accepted pair by itself; a person is asked. That is the honest shape — the value
+was never the mesh's to make.
+
+## How it is checked
+
+An inventory test accepts a value into a pair, reads it back twice unchanged with origin
+`accepted`, asserts `rotate` refuses it naming the remedy while a made pair still rotates;
+another changes a node's key and asserts the read is refused, then accepts again and reads.
+
+## References
+
+- [issue 070](../04-ISSUES/070-an-operator-cannot-deliver-a-pair-credential/00-report.md), [issue 069](../04-ISSUES/069-one-secret-provision-yields-one-value/00-report.md)
+- [ADR 0085](0085-a-secret-is-a-provision.md)
+- [`03-DESIGN/01-to-be/24-the-secrets-vault.md`](../03-DESIGN/01-to-be/24-the-secrets-vault.md)

02-DECISIONS/0093-a-fixture-that-runs-a-modules-runtime-carries-its-name.md

@@ -0,0 +1,53 @@
+---
+topic: checking it
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0089-a-bed-reads-the-catalogue-it-proves.md
+---
+
+# 93. A fixture that runs a module's runtime carries the module's name
+
+## Context
+
+[ADR 0089](0089-a-bed-reads-the-catalogue-it-proves.md) said a bed that needs less than a module
+declares is a mesh test, and carries a name of its own. Twelve beds were to be renamed on that
+basis ([issue 074](../04-ISSUES/074-a-mesh-test-wears-a-catalogue-modules-name/00-report.md)).
+Reading how a module's tools are reached showed why they cannot be: the account the mesh issues
+a module is scoped to `serve.<module>.*` from the manifest's name, and the runtime binds its tool
+queues under the name baked into its image. A fixture named otherwise but running the real
+runtime would be refused its own queues. The name is not a label; it is the tool namespace and
+the broker scope.
+
+## Considered Options
+
+1. **Rename anyway, and rebuild each runtime under the fixture's name.** Rejected: a runtime built
+   under a false name proves nothing about the module and costs a build per bed.
+2. **A fixture that runs a module's runtime carries the module's name, and therefore reads the
+   catalogue.** Adopted.
+
+## Decision
+
+A bed that runs a module's runtime installs the catalogue's manifest for that module and what it
+requires — the vault for a `secret`, the route module for a route — and proves the mechanism
+against the real module. A name of its own is for a fixture that runs no real runtime: a
+declaration-only stub, a bare upstream image.
+
+## Consequences
+
+The mechanism beds become module beds with a mechanism inside them, which is more than they
+were. What got harder: a bed that wanted a cut-down redis now raises the vault beside it; one that
+wanted a sidecar without its server raises the server. Each conversion is a lab run, and the beds
+still carrying a copy are declared with this reason until converted.
+
+## How it is checked
+
+The lab's inline-copy check refuses undeclared copies as before; the declared list's reason for
+these beds names this record. Two beds converted with the vault beside them ran green.
+
+## References
+
+- [issue 074](../04-ISSUES/074-a-mesh-test-wears-a-catalogue-modules-name/00-report.md)
+- [ADR 0089](0089-a-bed-reads-the-catalogue-it-proves.md), [ADR 0047](0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md)
+- [`03-DESIGN/01-to-be/01-end-to-end-testing.md`](../03-DESIGN/01-to-be/01-end-to-end-testing.md)

02-DECISIONS/0094-a-module-may-hold-several-secrets-from-one-provider.md

@@ -0,0 +1,63 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0085-a-secret-is-a-provision.md
+---
+
+# 94. A module may hold several secrets from one provider, each a pair of its own
+
+## Context
+
+[ADR 0085](0085-a-secret-is-a-provision.md) makes a module's own secret a provision: the module
+requires `secret` from the vault and reads the pair credential minted for that consumer↔vault
+pair. A pair has one credential, a module requires a provision once, and so a module received
+one value. Read against the catalogue, nine modules hold two or more secrets besides their
+broker account; seven of them hold genuinely independent values with independent lifetimes — a
+root certificate, its key and that key's password; an admin password beside an API token. None
+derives from another, so "one value, derivation the module's business" answers nothing
+([issue 069](../04-ISSUES/069-one-secret-provision-yields-one-value/00-report.md)).
+
+## Considered Options
+
+1. **One value per module; the module derives the rest.** Rejected: the values are independent.
+2. **Require the provision several times.** Rejected: `requires` is a list of names, and a
+   requirement is matched by name everywhere.
+3. **The `secrets` map names several files under local names, and each local name is a pair
+   credential of its own.** Adopted.
+
+## Decision
+
+A module's `secrets` entry for a requirement may be a path, as before, or an object of local
+names to paths. Each local name is its own pair credential, keyed on it beside the provision,
+the consumer node, the consumer module and the provider; its own file on the consumer, referred
+to as `${secret:<local name>}`; its own holder at the provider, named the consumer's identity
+with the local name after it; and rotated apart from the others. A local name may not be one of
+the module's own secrets or something it requires, so what a placeholder means is never
+ambiguous. The plain shape is unchanged, and every credential that exists is the one it was.
+
+The holder's suffix is not a login any backend checks — a secret is not a login — so the
+identity limit that binds a database role or an access key does not apply to it.
+
+## Consequences
+
+The ten modules that could not move onto the vault can. What got harder: `rotate secret` for a
+consumer rotates every local name it holds from that provider; rotating one of several is a
+finer command than the mesh has, and waits for a case that needs it.
+
+## How it is checked
+
+Manifest tests read both shapes, write them back, and refuse a colliding or unusable local
+name. A resolver test asserts two local names are two needs, two files with two credentials,
+and two holders at the provider. An inventory test asserts two local names are two rows, that
+rotating one leaves the other, and that the provider is told both. The vault bed installs a
+consumer that keeps two secrets and asserts two values delivered, two holders in the vault's
+ledger, and both rotated by one command.
+
+## References
+
+- [issue 069](../04-ISSUES/069-one-secret-provision-yields-one-value/00-report.md)
+- [ADR 0085](0085-a-secret-is-a-provision.md), [ADR 0049](0049-a-consumers-identity-fits-the-tightest-backend.md)
+- [`03-DESIGN/01-to-be/24-the-secrets-vault.md`](../03-DESIGN/01-to-be/24-the-secrets-vault.md)

02-DECISIONS/0095-the-control-plane-is-the-way-to-ask-a-module.md

@@ -0,0 +1,60 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md
+---
+
+# 95. The control plane is the way to ask a module
+
+## Context
+
+A module serves tools over the broker under an account scoped to what it emits, consumes and
+serves ([ADR 0047](0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md)). A
+tool call is a request and a reply: the caller creates a reply queue and publishes to the serving
+module's request key, and no module's scope grants either — nor should it, since a module that
+only publishes events has no business declaring queues. So a module could serve tools and nothing
+in the mesh could call them
+([issue 049](../04-ISSUES/049-a-module-can-serve-tools-and-nothing-can-call-them/00-report.md)):
+not an operator at a terminal, not an agent acting for one.
+
+## Considered Options
+
+1. **Calling is a grant**: a module declares it may be asked, and a consumer is issued an
+   account that may create a reply queue and publish to that module's request key. Deferred: a
+   module-to-module call is the only caller that resembles what the mesh mints today, and none
+   asks for one yet.
+2. **The control plane is the way in.** Adopted. It holds a connection that may already, so a
+   person or an agent asks through it, and every question passes one process — which is where
+   an audit of who asked what belongs.
+
+## Decision
+
+`mesh-controller ask <module> <tool> [json]` publishes the request on the tool exchange under
+`<module>.<tool>`, with a private reply queue bound under its own name, waits for the answer
+whose correlation matches, and prints it as the module gave it. A tool that answered with an
+error has answered: the answer is printed and the exit status says so. A module that never
+answers is said to have not answered, with where to look.
+
+A module declares nothing about being asked: serving a tool is being askable through the control
+plane. A module-to-module call, if one is wanted, is a grant like any other and a later decision.
+
+## Consequences
+
+Anything with the control plane in reach can ask any module anything it serves. What got
+harder: nothing outside the control plane can, and the control plane's connection is one more
+thing on the path of every question — a cost accepted for the audit it buys.
+
+## How it is checked
+
+A tools-only bed asks a served tool through the control plane and asserts an answer arrived —
+an error, since the lab has no upstream and no token, which is an answer where a timeout would
+not be.
+
+## References
+
+- [issue 049](../04-ISSUES/049-a-module-can-serve-tools-and-nothing-can-call-them/00-report.md)
+- [ADR 0047](0047-a-module-runs-its-code-as-its-own-process-with-its-own-account.md)
+- [`03-DESIGN/01-to-be/19-the-module-protocol.md`](../03-DESIGN/01-to-be/19-the-module-protocol.md)

02-DECISIONS/0096-an-upstream-image-is-copied-between-registries.md

@@ -0,0 +1,66 @@
+---
+topic: building it
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0006-the-substrate-and-the-control-plane.md
+---
+
+# 96. An upstream image is copied between registries, never through a machine's image store
+
+## Context
+
+A module may declare that an artifact is an image published elsewhere, to be copied into the
+mesh's own registry so machines fetch it by a digest this mesh assigned rather than by a name
+somebody else controls. The builder pulled it into the build machine's image store and pushed it
+under the mesh's name, and the push was refused: a published image is an index over several
+architectures, the runtime's store keeps the index, and pushing one platform out of it fails
+however the platform is asked for
+([issue 046](../04-ISSUES/046-an-upstream-image-cannot-be-mirrored-into-the-mesh/00-report.md)).
+Every variant of pull-then-push was tried and failed the same way.
+
+## Considered Options
+
+1. **Resolve the index to one platform and push that.** Tried, reverted: it did not make the
+   push work, and a workaround for a store's behaviour is a thing nobody removes later.
+2. **Tooling that copies between registries**, installed on the build machine. Rejected: one
+   more thing the builder's image carries, for a protocol the builder already speaks for blobs.
+3. **Copy over the registry API**, in the builder. Adopted.
+
+## Decision
+
+The builder copies an upstream image between registries and never through a machine's image
+store: it reads the index and every manifest it names, moves each blob by digest into the mesh's
+registry — skipping what is already there, since blobs are content-named — puts the manifests
+and then the index under the module's repository, and pins the index's digest. Public images are
+read with the anonymous bearer token the registry hands out on challenge, which is how the
+public hub and the others the catalogue names serve them. The mesh mirrors the whole index, so
+what a machine fetches is the image for its own architecture; that every machine on one mesh is
+the same architecture is an assumption this mesh makes and had not written down until now.
+
+Genesis has no registry to copy into and keeps the pull: the image stays in the first machine's
+store, named by its own id, as every artifact does before there is anywhere to publish.
+
+## Consequences
+
+An upstream artifact builds. What got harder: the builder now holds a registry client of its
+own, some two hundred lines, where a runtime command used to do; and a private upstream that
+demands a credential is refused, since the copy is anonymous by design.
+
+## How it is checked
+
+A test raises a fake upstream registry serving an index over two platforms behind a bearer
+challenge, and a fake mesh registry that records what arrives: every blob of both platforms
+arrives once, two manifests and the index are put under their digests, the reference returned
+pins the index under the module's repository, and a second copy uploads nothing. A reference
+test reads names the way a runtime does. *Proven against the real thing the same day:* the genesis
+bed built the tool runtime through the mesh's builder with its node base copied out of the public
+hub into the mesh's registry by this code — after one finding the fake could not give: the builder
+ran on the default bridge, where loopback is not the machine, and now runs on the host network.
+
+## References
+
+- [issue 046](../04-ISSUES/046-an-upstream-image-cannot-be-mirrored-into-the-mesh/00-report.md)
+- [ADR 0006](0006-the-substrate-and-the-control-plane.md)
+- [`03-DESIGN/01-to-be/18-building-a-module.md`](../03-DESIGN/01-to-be/18-building-a-module.md)

02-DECISIONS/0097-a-vendor-image-is-a-declared-build-input.md

@@ -0,0 +1,65 @@
+---
+topic: building it
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0096-an-upstream-image-is-copied-between-registries.md
+---
+
+# 97. A vendor image is a declared build input, and a recipe fetches nothing undeclared
+
+## Context
+
+Three modules could not be built by the mesh's builder because their recipes reached for what
+no manifest named: a public package, or a binary copied out of a public image
+([issue 064](../04-ISSUES/064-a-mesh-build-cannot-fetch-a-modules-external-dependencies/00-report.md)).
+A module already names the bases it stands on — another module's artifact, by name — and the
+builder answers with what the mesh holds; a vendor's image had no such declaration, so a recipe
+named it directly and the build worked when the public registry answered, which is sometimes.
+
+## Considered Options
+
+1. **Let the build environment reach public registries.** Rejected: a build that fetches from
+   somebody else's registry on its own is one the mesh cannot rebuild the same way twice.
+2. **A vendor image is a base like any other**, declared under `build.on` with the argument the
+   recipe reads it from, pinned by digest, copied into the mesh's registry before the build
+   ([ADR 0096](0096-an-upstream-image-is-copied-between-registries.md)). Adopted.
+
+## Decision
+
+A build's `on` entry is either a module's artifact or an image published elsewhere, pinned by
+digest, read from one build argument. Before the build the image is copied into the mesh's
+registry under the module's repository and the recipe is handed the copy; genesis, with no
+registry, pulls it into the first machine's store. A recipe whose `COPY --from` names a registry
+image the manifest did not declare is refused before the build, naming the image and the remedy;
+its own stages, declared arguments and `scratch` are not fetches. An unpinned vendor image is
+refused: a tag is what somebody else can move.
+
+A recipe whose `FROM` names an undeclared base was at first said, not refused: the mesh's own
+images — the control plane's, the tool runtime's, the route proxy's — started from a public base
+and declared none, and refusing those refuses genesis. *Amended the same day:* those three declare
+their bases now, and an undeclared `FROM` is refused like an undeclared copy. The builder's own
+image and the examples are built by `make`, not by the mesh, and take arguments with defaults.
+
+The package half of the issue is not decided here: the mesh's package registry already proxies
+the public one, and the failure the report saw has to be run again to be placed.
+
+## Consequences
+
+A module's build inputs are all in its manifest, and every one of them is something the mesh
+holds a copy of. What got harder: a recipe that used to name a base image on its first line now
+names an argument, and the manifest names the image.
+
+## How it is checked
+
+A builder test declares a pinned vendor image, asserts it is copied under the module's
+repository and handed to the recipe as the argument, and asserts an unpinned one is refused. A
+recipe test asserts an undeclared `FROM`, an undeclared `COPY --from` and an undeclared argument
+are named, and that stages, declared arguments and `scratch` are not.
+
+## References
+
+- [issue 064](../04-ISSUES/064-a-mesh-build-cannot-fetch-a-modules-external-dependencies/00-report.md)
+- [ADR 0096](0096-an-upstream-image-is-copied-between-registries.md)
+- [`03-DESIGN/01-to-be/18-building-a-module.md`](../03-DESIGN/01-to-be/18-building-a-module.md)

02-DECISIONS/0098-a-fact-a-provider-makes-at-first-start-is-fetched-from-it.md

@@ -0,0 +1,61 @@
+---
+topic: the tiers
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0085-a-secret-is-a-provision.md
+---
+
+# 98. A fact a provider makes at first start is fetched from it, not carried in its manifest
+
+## Context
+
+The catalogue's certificate authority declared its root certificate, its root key and that key's
+password as its own secrets, and told the container to initialise from them. The mesh mints an
+own secret nobody delivers as random bytes, and random bytes are not a certificate: issued, the
+authority could not start; only an operator hand-making its root could raise it
+([issue 076](../04-ISSUES/076-a-served-fact-made-at-first-start-cannot-be-served/00-report.md)).
+The authority can make its own root at first start. What it could not do then was tell the mesh
+what that root is: a consumer was given `${bound:acme-ca:root}` from the provider's `serves`,
+which is written in the manifest before anything runs.
+
+## Considered Options
+
+1. **A secret the module makes**, with the mesh taking custody once the file exists. Rejected
+   for now: a node would have to send a value up to the mesh, which no channel does today, and
+   a root key is the one thing the mesh has no reason to hold.
+2. **A served fact the provider contributes at run time.** Rejected for now: the same new
+   channel, for a fact that is not secret at all.
+3. **The consumer fetches it from the provider**, over the mesh network, through a gate before
+   the thing that needs it starts. Adopted.
+
+## Decision
+
+A provider's `serves` names where a fact made at first start can be fetched — the authority
+serves its root at a path beside its ACME directory — and a consumer fetches it in a `run-once`
+step declared before the resource that needs it, from the provider's bound address. The mesh
+network is where the fetch happens, which is what makes fetching without a prior trust
+acceptable: it is the network the mesh itself authenticates. The mesh mints only what it can
+make: the authority's password. The root key stays where it was made.
+
+## Consequences
+
+The catalogue's authority starts, and the proxy that requires it trusts what it fetched. What
+got harder: a consumer of such a fact carries one more resource, the gate that fetches it, and
+a fact that changes after first start is refetched only when the declaration changes.
+
+## How it is checked
+
+The route-forwarding bed installs the authority, the proxy and a consumer from the catalogue and
+asserts a routed name is served through the proxy. The proxy refuses to start on a bundle that is
+not a certificate, so the name being served proves the gate fetched one; the gate itself refuses
+a body that is not a certificate. That the proxy obtains a certificate from this authority through
+that root is the certificate bed's proof, against the same authority with the same proxy. The
+catalogue-wide manifest test parses both manifests.
+
+## References
+
+- [issue 076](../04-ISSUES/076-a-served-fact-made-at-first-start-cannot-be-served/00-report.md)
+- [ADR 0053](0053-a-step-that-runs-on-a-schedule.md), [ADR 0085](0085-a-secret-is-a-provision.md)
+- [`03-DESIGN/01-to-be/08-connectivity.md`](../03-DESIGN/01-to-be/08-connectivity.md)

02-DECISIONS/0099-a-step-that-runs-once-names-what-it-reads.md

@@ -0,0 +1,94 @@
+---
+topic: what runs on it
+status: accepted
+date: 2026-09-21
+deciders: jochen
+reconstructed: false
+extends: 0052-a-step-that-runs-once-before-a-container.md
+---
+
+# 99. A step that runs once names what it reads, and runs again when it changed
+
+## Context
+
+[ADR 0098](0098-a-fact-a-provider-makes-at-first-start-is-fetched-from-it.md) has a consumer
+fetch a fact its provider made at first start through a run-once step: the route proxy fetches
+the certificate authority's root before it starts. A run-once step runs once per declaration
+([ADR 0052](0052-a-step-that-runs-once-before-a-container.md)): its marker is the digest of its
+own declaration, and a re-apply that finds the marker does nothing.
+
+The provider can move. When the authority is assigned to another node it makes a new root there,
+and the mesh rewrites the consumer's binding file with the new address — but the step's own
+declaration has not changed, so the step does not run again, the proxy keeps the old root, and it
+refuses every certificate the new authority issues
+([issue 077](../04-ISSUES/077-a-fact-fetched-at-first-start-is-fetched-once/00-report.md)). A
+restart trigger was the natural remedy and was refused on a run-once step, on the ground that a
+step does not stay running to be restarted.
+
+Two things the host already does point at the answer. What a container reads is part of what it
+is: a container's digest includes the digest of every resource it names under `restart-on`, so a
+rewritten file it reads is a changed container
+([issue 045](../04-ISSUES/045-a-container-keeps-the-values-it-started-with/00-report.md)).
+And a run-once step's marker *is* its digest. Nothing new is needed for the step to run again when
+what it reads changed; only the refusal stands in the way.
+
+## Decision
+
+**A run-once step may name what it reads under `restart-on`. For a step the word means *run
+again*: when a named resource changed in this apply, the step's digest has moved, its marker no
+longer matches, and it runs again — gating what follows, as it did the first time.** Nothing
+about the marker changes; the refusal of the pair is lifted, in the control plane and on the host.
+
+**The container that consumes what a step made names the step.** A step that ran counts as a
+change, so a service that names it under `restart-on` is recreated after it, holding what the step
+fetched. Without this the step fetches a new root and the service keeps serving with the old one.
+
+The route proxy's gate names the binding file it reads; the proxy's server names the gate and the
+binding. When the authority moves, the binding is rewritten, the gate fetches the new root, and the
+server is recreated with it — in one apply.
+
+## Considered Options
+
+1. **A provider epoch in the binding — the mesh raises a number when a provider is re-issued or
+   moved, and the consumer's file carries it.** Rejected: the binding already changes when the
+   provider moves (its address does), and a re-issue does not change what the authority serves —
+   its state persists. An epoch would be a second signal for a change the file already shows.
+2. **The step runs before every start of the service, with no marker.** Rejected: every reconcile
+   would run it, and a step that runs on every apply reads as a change on every apply, so the
+   service naming it would be recreated every few minutes.
+3. **The proxy fetches the root itself, at start.** Rejected as the general answer: it fixes the
+   proxy and leaves the next consumer of a fact made at first start to fix itself. The step is the
+   general shape ([ADR 0098](0098-a-fact-a-provider-makes-at-first-start-is-fetched-from-it.md)).
+4. **Lift the refusal and read `restart-on` as *again* on a step.** Adopted: it is what the
+   digest already does, and it needs no new word.
+
+## Consequences
+
+A fact fetched at first start follows its provider when the provider moves. What is not covered:
+a provider whose state is wiped behind the mesh's back, on the same node, makes a new fact that
+nothing the mesh knows reflects. That is not a change the mesh can see, and it is not claimed.
+
+The one contradiction the refusal named is real and is now a documented reading: on a running
+container `restart-on` means recreate, on a step it means run again. Both are "this must reflect
+what it reads". This record is about a run-once *container*, the step ADR 0052 defined. A
+run-once process is a different shape whose marker does not carry what it reads; it is not
+covered here.
+
+## How it is checked
+
+- mesh-host: a unit test declares a run-once step naming a file, records its marker against the
+  file's old content, applies with the new content and asserts the step ran; applies again with
+  nothing changed and asserts it did not. A second test declares a container naming a run-once
+  step and asserts the container is recreated after the step ran, with the step as the stated
+  reason.
+- mesh-controller: the manifest parser accepts a run-once step with `restart-on`; the
+  catalogue-wide manifest test parses the route proxy's manifest, whose gate and server name what
+  they read.
+- The route-forwarding bed still passes with the host that accepts the pair. No bed moves the
+  authority: the mechanism is proven by the unit tests, the declaration by the manifest test.
+
+## References
+
+- [issue 077](../04-ISSUES/077-a-fact-fetched-at-first-start-is-fetched-once/00-report.md)
+- [ADR 0052](0052-a-step-that-runs-once-before-a-container.md), [ADR 0053](0053-a-step-that-runs-on-a-schedule.md), [ADR 0098](0098-a-fact-a-provider-makes-at-first-start-is-fetched-from-it.md)
+- [`03-DESIGN/01-to-be/08-connectivity.md`](../03-DESIGN/01-to-be/08-connectivity.md), [`03-DESIGN/01-to-be/20-writing-a-module.md`](../03-DESIGN/01-to-be/20-writing-a-module.md)

02-DECISIONS/0100-a-node-in-use-is-adopted-before-it-is-converged.md

@@ -0,0 +1,221 @@
+---
+topic: the mesh
+status: accepted
+date: 2026-09-22
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0078-the-store-and-broker-are-modules.md
+---
+
+# 100. A node in use is adopted before it is converged
+
+## Context
+
+The mesh replaces a predecessor mesh that is running, on the same machines, with the services
+people use. The control-node is the machine that already carries the predecessor's broker and
+build pipeline. The operator proposed the migration's shape: stop the predecessor's control on a
+machine, bring the mesh up there in an adoption mode that keeps the machine's configuration in
+force, migrate its modules one at a time, move to the next machine, and flip adoption off when every
+machine is done.
+
+Measured on the control-node ([research 012,
+*migrating a node that is in use*](../01-RESEARCH/012-the-minimum-viable-node/migrating-a-node-in-use.md)):
+60 predecessor containers; the predecessor's control in user-level daemons separate from its
+services; its firewall active, with 54 incoming and 52 forwarding rules allowing each served port
+explicitly; 12 files its configuration sync writes, 3 of them system files.
+
+Four things in the mesh as it stands break that shape:
+
+1. **The base ruleset closes the machine.** Genesis loads a drop-by-default table
+   ([ADR 0088](0088-the-foundation-filters-before-anything-listens.md)). Every base chain at a hook
+   runs in priority order; an accept ends only its own chain and a drop in any is final — whether
+   the other firewall's chains are nftables or legacy iptables. So the predecessor's allowed ports
+   would close at genesis.
+2. **The host replaces what it finds.** A declared file is written whatever is at its path, except
+   a file declared create-once. A declared container replaces a running one of the same name. So
+   assigning a module the predecessor also runs replaces the predecessor's service and its files at
+   once.
+3. **Some foundation ports are held.** The foundation's store and bus ports are free — the
+   predecessor publishes its own elsewhere — but the registry's port and the broker's management
+   port are held, and on a control-node that is the private network's hub, so is the private
+   network's port. The foundation's ports are fixed in the installer's bundle and the catalogue's
+   manifests, so a collision surfaces as a container that fails to bind, and a port changed at
+   genesis would be changed back when the foundation is adopted as modules.
+4. **Published container ports are forwarded, not received.** The foundation publishes its ports
+   on every interface; a predecessor firewall that filters only incoming traffic never sees them.
+   The base ruleset is what keeps the store unreachable from outside, and it is the thing that
+   cannot be loaded.
+
+The mesh already adopts in one place: the foundation's store and broker are taken over in place as
+modules, keyed on the container that is already running
+([ADR 0078](0078-the-store-and-broker-are-modules.md)). The research this record draws on already
+settled the conflict rule for adoption: *on conflict, what is on the machine stays*.
+
+## Considered Options
+
+1. **Cut each machine over in one go** — stop the predecessor's store, broker, registry and proxy,
+   raise the foundation in their place. Rejected: every predecessor service on the machine is down
+   until it has migrated, the predecessor's other machines lose their broker, and the rollback is
+   restarting the predecessor — a recovery, not a step.
+2. **Put the control-node on a separate machine.** Rejected: the control-node is decided.
+3. **Converge on joining, as today.** Rejected: the base ruleset closes the machine at genesis, and
+   the predecessor's services and files are replaced as soon as any module naming them is assigned.
+4. **Only make the foundation's ports configurable.** Rejected as insufficient: it answers the bind
+   collisions and neither the firewall nor the files.
+5. **Treat assigning a module as migrating it.** Considered and rejected on review: it makes the
+   rule that keeps found files never fire — the host only ever sees files of assigned modules — and
+   it makes an assignment on an adopted node an outage rather than a preparation.
+6. **Adoption as a mode per node; each module taken explicitly; the node converged by an explicit
+   flip.** Adopted.
+
+## Decision
+
+**A node is adopted or converged, and the controller records which.** The operator says so: at
+genesis for the control-node, and in the enrolment token for the others. The controller is
+authoritative, and every declaration it sends says whether the node is adopted and which modules
+have been taken on it. A node stays adopted until the operator converges it. An adopted node is
+said to be adopted wherever the mesh reports a node's state.
+
+**A converged genesis refuses a machine in use.** A machine is in use when a container is running
+on it, or a port is listening on an address other than loopback that is not ssh's. Raised without
+saying adopted on such a machine, genesis refuses and names every container and listener it
+counted — a forgotten flag must not close a working machine.
+
+**Before a node is adopted, its predecessor's control is stopped by the operator** — the daemons
+that write its configuration. Its services keep running on what they have.
+
+**Found means present with no record.** A file at a declared path, or a container at a declared
+name, that the host's store has no record of writing is *found*. A file the host wrote in an
+earlier life of the node is not found; its record says so.
+
+**On an adopted node, what is found is kept until its module is taken.** The host keeps a found
+file and a found container as they are, records the file's original content before anything else
+happens to it, and reports each as held. Assigning a module on an adopted node prepares it: what
+the module declares that is not found is created; what is found is held. **Taking a module** on a
+node is its cutover — the operator's act, done when that module's data has moved — and from then
+on the module's resources converge on that node like any other. What is held is never removed,
+even when its module is unassigned, and a held file or container that changes while held — a file
+rewritten, a container stopped or replaced — is reported as changed by something else, not reverted
+or restarted: that is how a predecessor still writing is caught.
+
+**The firewall found on the machine stays in force.** The mesh loads no table on an adopted node
+that drops by default or holds an accept — neither genesis's base ruleset nor the filter module's
+derived one. What the mesh needs reachable is declared as **openings**: a resource that says a port
+is reachable, from where, on the incoming path or the forwarded path — a published container port is
+forwarded. The controller derives them from the same inputs as the filter, each from where the
+filter would admit it: the `listens` of the modules assigned there, the private network's hub port
+and the bus and the registry from anywhere, the store's port and the broker's management port from
+the private network. The host
+converges an opening through the found firewall in that firewall's own terms, marks it as the
+mesh's, and removes only what it marked; it re-checks each opening on every reconcile, so a reload
+or a reboot of the found firewall does not lose it for longer than one reconcile. An opening is a
+state, not a command, which is what lets it travel over the link. The host reports which firewall
+it found. A machine with no firewall needs no openings; a machine with a kind no host speaks is
+refused adoption.
+
+**The mesh guards its own ports itself, in a table of its own that only refuses.** It passes
+everything by default and holds nothing but refusals, and the two ports it refuses are the
+foundation's own, checked free at genesis, so it cannot close anything the machine serves; it is
+the mesh's, so the found firewall reloading does not touch it. It refuses the store's port and the
+broker's management port except from the private network and from the machine itself — its
+loopback and the container runtime's own networks, known by the interface a packet arrives on and
+never by its source address alone — at the prerouting hook, ahead of the runtime's
+destination translation, so it matches the port the packet was sent to, for both address
+families. The bus and the registry stay reachable from anywhere, as a node
+enrols over the bus and pulls from the registry before it has a private-network address
+([ADR 0088](0088-the-foundation-filters-before-anything-listens.md)); so does the private network's
+hub port. The store is unreachable from outside whatever the found firewall does, and on a machine
+with none.
+
+**The foundation's ports are the node's.** Every port the foundation binds is an input to genesis,
+checked free before anything is raised, refused with the name of what holds it. The ports given
+become that node's settings for the foundation's modules — the catalogue's numbers are only their
+defaults — and every place that uses them reads them from there: the modules' containers, the
+filter, the base ruleset, the private network's endpoint, and the addresses consumers are given.
+The private network's address range must not overlap a tunnel the predecessor still runs; genesis
+checks that too.
+
+**Converging a node is one act, previewed.** It refuses while an assigned module still holds a
+found container: each service is taken on its own, when its data has moved, never by the flip. The
+preview lists what is reachable on the machine now — every listening socket and every published
+container port — and for each whether an assigned module declares it or it will close, and every
+module the flip will take, with the held files each will replace. The flip then takes those modules, loads the
+mesh's derived filter in place of its refusal-only table, and retires the found firewall by
+disabling it, never by flushing: the container runtime's rules and the found firewall's own
+configuration stay on disk. Returning a converged node to adopted unloads the derived filter,
+restores the refusal-only table, enables the found firewall again and converges the openings
+through it once more; what was taken stays taken. A
+node converges when its migration is done; the mesh is migrated when every node has converged.
+
+**The order is the operator's:** the control-node first, adopted, its modules assigned and taken
+one at a time; then each other machine, adopted, migrated, converged in turn. The predecessor's
+pipeline runs on the control-node, so its updates stop for every machine while the migration runs;
+that is accepted.
+
+## Consequences
+
+Each step says what it changes before it changes it. Adopting a node changes nothing that serves;
+assigning a module adds what is not there; taking a module replaces one service; the flip replaces
+the firewall, after naming every port it will close. Two steps are not undone by the mesh: taking a
+module replaces the predecessor's container, and the kept original of a file is recorded but not
+yet restored by any act of the mesh
+([research 012](../01-RESEARCH/012-the-minimum-viable-node/00-overview.md) leaves where it lives
+open).
+
+What got harder:
+
+- **The mesh must speak a firewall it did not install**, on both the incoming and the forwarded
+  path. One kind is found on the machines measured; another is refused until a host speaks it. The
+  mesh's own guard does not depend on it: that table is the mesh's.
+- **The host gains a guard it did not have** — keep what you found — and its report must say
+  which files and containers it holds, or an adopted node reads as converged.
+- **The declaration gains a node's mode and its taken modules**, and a resource, the opening.
+- **Genesis grows inputs, and they outlive genesis.** The foundation's ports stop being constants;
+  every reader of them reads the node's settings.
+- **A node can sit adopted indefinitely.** Nothing forces the flip; the mesh's status says which
+  nodes are adopted, so one left behind is visible.
+- **This narrows [ADR 0088](0088-the-foundation-filters-before-anything-listens.md) for adopted
+  nodes**: the base ruleset is not loaded on a node raised adopted, and its duty — the store never
+  reachable from outside — passes to a table of the mesh's that only refuses.
+
+## How it is checked
+
+A lab bed prepares a machine the way the predecessor leaves one: its firewall allowing a served
+port and denying the rest, a service container listening on that port under a name a catalogue
+module also uses, a file at a path that module declares, a stand-in for the predecessor's control
+that would rewrite that file, and a container holding the registry's port. Then:
+
+- **Genesis converged** on it refuses and names every container and listener it counted.
+- **Genesis adopted, with the registry's port held**, refuses and names the holder; with another
+  port given, the foundation comes up — and adopting the foundation as modules leaves it on that
+  port.
+- **Nothing that serves changed**: the service is reachable from a second machine, the file is byte
+  for byte what it was, and the found firewall's rules differ only by rules marked as the mesh's.
+- **The store is unreachable from outside** — probed from a machine off the private network, and
+  again after the found firewall is reloaded — and reachable over it and from a container on the
+  node itself; the bus is reachable from a machine that has not yet enrolled.
+- **The mesh works through the found firewall, and keeps working after it is reloaded and after the
+  machine reboots**: the second machine enrols, and the openings are there again.
+- **A predecessor still writing is caught**: with the stand-in left running, the held file's change
+  is reported and not reverted.
+- **Assigning prepares, taking cuts over**: the module assigned holds the found container and file;
+  taken, it replaces them and its port is opened.
+- **Converging previews, then changes**: it refuses while the service's module holds its found
+  container; once that module is taken, the preview names the service's port and a published port no
+  firewall rule mentions, and the modules it will take; after the flip the mesh's derived filter is
+  loaded, the found firewall is disabled with its configuration still on disk, the declared port is
+  open and the undeclared one closed. Returned to adopted, the found firewall is enabled again and
+  the derived filter is gone.
+
+Unit tests hold the host to keeping a found file and container on an adopted node, converging them
+once taken, never removing what it holds, and reporting a held file or container that changed;
+genesis to refusing a held port and a converged raise on a machine in use; the controller to
+carrying the mode and the taken modules in every declaration, deriving the openings, and refusing
+a flip while a found container is held.
+
+## References
+
+- [research 012 — the minimum viable node, and adopting what is already there](../01-RESEARCH/012-the-minimum-viable-node/00-overview.md),
+  and its document [*migrating a node that is in use*](../01-RESEARCH/012-the-minimum-viable-node/migrating-a-node-in-use.md)
+- [ADR 0005](0005-the-node-host.md), [ADR 0011](0011-managed-files-are-generated-never-edited.md),
+  [ADR 0078](0078-the-store-and-broker-are-modules.md), [ADR 0088](0088-the-foundation-filters-before-anything-listens.md)

02-DECISIONS/0101-a-machines-own-resolver-does-not-make-it-in-use.md

@@ -0,0 +1,70 @@
+---
+topic: the mesh
+status: accepted
+date: 2026-09-22
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0100-a-node-in-use-is-adopted-before-it-is-converged.md
+---
+
+# 101. A machine's own resolver does not make it in use
+
+## Context
+
+[ADR 0100](0100-a-node-in-use-is-adopted-before-it-is-converged.md) has a converged genesis
+refuse a machine in use. It defines *in use* as a container running, or a port listening on an
+address other than loopback that is not ssh's. That definition was written before anything was
+measured.
+
+Measured on a freshly installed lab machine, running nothing but its operating system:
+
+| Listening | Held by |
+|---|---|
+| TCP and UDP on every address, the link-local name resolution port | the system's name resolver |
+| UDP on every address, the multicast name resolution port | the system's name resolver |
+| UDP on a link-local address, the address-configuration client port | the system's network manager |
+| TCP and UDP on loopback | the resolver's stub and the container runtime |
+
+By ADR 0100's words, the resolver's TCP listener on every address makes **every** freshly
+installed machine a machine in use. A converged genesis would refuse them all, and `--adopted`
+would become the only way to raise anything. The refusal exists to catch a forgotten flag on a
+working machine. Refusing an empty one defeats it, and teaches operators to pass the flag by
+habit.
+
+## Considered Options
+
+1. **Keep the words.** Rejected: every fresh machine is refused.
+2. **Count TCP only, ignore UDP.** Rejected: the resolver listens on TCP too, and a machine that
+   serves over UDP alone, a resolver or a tunnel, is in use.
+3. **Ignore listeners held by the operating system's own network daemons**, a short named list,
+   on both protocols. Adopted.
+
+## Decision
+
+**A listener held by one of the operating system's own network daemons does not make a machine
+in use.** The daemons are the ones the measurement found: the name resolver and the network
+manager, named in the installer's code beside that measurement. Everything else in ADR 0100's definition stands: a running container, or any other
+listener on an address other than loopback that is not ssh's, makes the machine in use, and
+genesis still names every one it counted.
+
+A daemon is added to the list only with a measurement of a fresh machine that holds it.
+
+## Consequences
+
+- A converged genesis on a fresh machine goes ahead, as it did before ADR 0100.
+- A machine whose resolver is also serving other machines is not counted as in use by its
+  resolver alone. It is one of the daemons that serves nobody on a fresh machine, and the one
+  kind of service this lets through.
+- The list is code, not configuration, so it changes by review.
+
+## How it is checked
+
+A unit test in the installer feeds the listeners captured from the fresh machine, as the
+listening-socket tool printed them, and asserts the machine is not in use. The existing tests
+still assert that a serving machine is in use and that every container and listener is named.
+The adoption lab bed raises a converged genesis on a fresh machine and asserts it is not refused.
+
+## References
+
+- [ADR 0100](0100-a-node-in-use-is-adopted-before-it-is-converged.md)
+- [research 012, *migrating a node that is in use*](../01-RESEARCH/012-the-minimum-viable-node/migrating-a-node-in-use.md)

02-DECISIONS/0102-the-mesh-writes-into-a-shared-file-never-over-it.md

@@ -0,0 +1,93 @@
+---
+topic: the mesh
+status: accepted
+date: 2026-09-22
+deciders: jochen
+reconstructed: false
+extends: 02-DECISIONS/0100-a-node-in-use-is-adopted-before-it-is-converged.md
+---
+
+# 102. The mesh writes into a shared file, never over it
+
+## Context
+
+[Issue 084](../04-ISSUES/084-taking-networking-on-an-adopted-node-restarts-every-container/00-report.md)
+found that the networking module declares the container runtime's configuration file to state
+one fact in it: the mesh's registry is trusted over the private network. Reading the code
+closer showed it worse than reported. The controller merges an operator's settings into the
+module's own content, but the host writes the result **whole**. Whatever the machine had in that
+file is replaced, including where the runtime keeps its data. On a machine in use, that is every
+image and container gone from the runtime's view at its next start. The runtime's service is then
+restarted, which stops every container on the machine.
+
+Measured on a lab machine: the runtime takes a new list of trusted registries on a reload, with
+no restart, and a running container with no restart policy keeps running through it. The
+runtime's log says it reloaded its configuration, and the registry reads as trusted afterwards.
+
+Under [ADR 0100](0100-a-node-in-use-is-adopted-before-it-is-converged.md), the file is found
+on an adopted node and held until networking is taken. Holding it is safe, but it means an
+adopted machine cannot pull the mesh's images until then, and taking networking would restart
+the runtime.
+
+## Considered Options
+
+1. **Keep writing the whole file; take networking as a cutover of its own.** Rejected: it
+   still replaces the machine's settings, and the cutover stops everything on the machine.
+2. **A drop-in the runtime reads beside its main file.** Rejected: the runtime has no such
+   directory for its daemon settings.
+3. **Write into the file: set the mesh's keys, keep the rest; reload, don't restart.** Adopted.
+
+## Decision
+
+**A file the mesh shares with software it did not install is written into, never over.** A
+file resource may say it is written *into* a structured file. The host then reads what is there,
+sets only the keys the mesh declares, keeps every other key as it found it, and records what each
+of its keys held before. **A list is added to, never replaced**: where the machine already has a
+list under a key the mesh declares, the mesh's members are added to it and the host records
+exactly which members it added — setting the key would replace the operator's own list, the harm
+this record exists to prevent. Undeclared later, each key goes back to what it held and each
+added member is taken out again, and a file the mesh created is removed only if nothing but its
+own keys is left. A file written into replaces nothing, so on an adopted node it is never held:
+it is written whether or not its module has been taken.
+
+**Whatever the host writes over without a record of it, it keeps first.** On any node, adopted
+or converged, before the host writes a file over one it has no record of making, it keeps the
+original once and says where; if it cannot keep it, it does not write. A file the mesh takes over
+is then never lost, whatever put it there.
+
+**A service that re-reads its configuration on a reload is reloaded, not restarted.** A service
+resource may name what it must be *reloaded* on, beside what it must be restarted on. The
+container runtime is reloaded for the registry's trust.
+
+**The networking module writes the runtime's trust into its file and reloads it.** An adopted
+node therefore trusts the mesh's registry as soon as it is on the private network, and taking
+networking no longer touches the runtime. The hosts file networking writes is still written whole
+and stays held until networking is taken; a converge preview names it among the files it replaces.
+
+## Consequences
+
+- The runtime's file on a machine in use keeps its data directory, its logging settings and
+  everything else the predecessor set.
+- A host that does not know *into* or *reload-on* refuses a declaration carrying them, so hosts
+  are upgraded before the controller that emits them — the same order ADR 0100 needs.
+- One shape more for every host: a file written into a structured document. Only JSON is spoken;
+  another format is refused until written.
+- The hosts file remains a whole file. Writing a marked block into it is the same idea for a text
+  file and is not decided here.
+
+## How it is checked
+
+Unit tests hold the host to setting only the declared keys and keeping the rest, restoring each
+key and removing only a file it created when the resource is undeclared, adding to a list and
+removing only the members it added, keeping the original of a file it writes over without a
+record, refusing a file that is not a JSON object rather than overwriting it, never holding a file written into on an adopted
+node, and reloading rather than restarting a service whose reload-on resource changed. A test in
+the controller holds the networking module to declaring the runtime's file written into and the
+runtime reloaded. The adoption lab bed gives the machine a runtime file with a setting of its own
+and asserts it survives adoption with the registry trusted added, and that a container without a
+restart policy is still running afterwards.
+
+## References
+
+- [Issue 084](../04-ISSUES/084-taking-networking-on-an-adopted-node-restarts-every-container/00-report.md)
+- [ADR 0082](0082-the-registry-is-reached-by-name-and-trusted-by-the-overlay.md), [ADR 0100](0100-a-node-in-use-is-adopted-before-it-is-converged.md)