Fixes issue 381 (a merge asked for a named head merged a newer head instead).
What happened, read from the forge
mesh-controller #197's timeline: the checks passed at 065cfa0d at 11:59 UTC (the head the review approved); the branch was force-pushed to 3ced96f at 12:14; the checks failed at an intermediate d0c9b9b4 at 12:17 and passed at 3ced96f at 12:22; the merge took 3ced96f at 12:26. Twelve minutes passed between the push and the merge, so this was not the forge's record of the head lagging the branch.
The report says the merge was asked for with head_commit_id 065cfa0d. It could not have been: gitea_merge_pull_request takes no argument for a head. It reads the pull request's head at the moment of the merge, checks that head's statuses, and hands that same head to the forge. So it kept its own promise — only a checked head is merged — but nothing could tell it which head the review had approved, and a commit nobody reviewed landed.
What changes
gitea_merge_pull_request takes head: the commit the caller reviewed, as a sha or at least its first 7 characters, compared case-insensitively. It is refused with nothing sent when head is missing, is not a commit, or is not the pull request's head; the refusal names the head the pull request has now. The forge is handed the full head as head_commit_id, so a push landing between this read and the merge is the forge's to refuse. The answer now carries head, and reads the pull request back after the merge: if the forge merged a different head anyway, the answer says so loudly (landed, not_the_head_named) rather than reporting the merge as the one asked for.
head is required, not optional — a decision I made here, for the reviewer to judge. An optional guard is the shape that let this happen. I found no code in any of the mesh's repositories (mesh-catalog, mesh-controller, mesh-tools, mesh-host, mesh-sdk, mesh-media-catalog) and no hq document that calls this tool; only agent sessions do, through the console. A session that leaves head out is told what to name.
Not verified
Whether Gitea 1.27.3 (the forge's version) actually refuses a merge whose head_commit_id differs from the branch's head at merge time. The tool already sent it before this change, and nothing here depends on it: the tool's own comparison is the guard; the forge's is the backstop for the read-to-merge window, and the read-back reports if it does not hold.
The tests were not run on this machine. The module's local dependencies are stale (an older @novox/mesh-sdk without the events, tools and provisioner subpaths, and no undici-types), so it does not type-check here for reasons outside this change. mesh/repo-check's TypeScript part, in the toolchain on the build node, is the first run of the new tests.
Tests (test/guard.test.ts)
The fake forge's head is now a full 40-character sha a test can move, and the merge marks the pull request merged and can report a different head after it. Existing merge tests name the head they were reviewed at. New: no head, a blank head, a branch name and a 5-character prefix are refused; a head that moved after review is refused for both the full sha and its 7-character prefix, naming the new head, with nothing sent; a 7-character upper-case prefix merges and hands the forge the full sha; a forge that merges another head is reported as such.
Fixes issue 381 (a merge asked for a named head merged a newer head instead).
## What happened, read from the forge
mesh-controller #197's timeline: the checks passed at **065cfa0d** at 11:59 UTC (the head the review approved); the branch was force-pushed to **3ced96f** at 12:14; the checks failed at an intermediate d0c9b9b4 at 12:17 and passed at 3ced96f at 12:22; the merge took **3ced96f** at 12:26. Twelve minutes passed between the push and the merge, so this was not the forge's record of the head lagging the branch.
The report says the merge was asked for with head_commit_id 065cfa0d. It could not have been: `gitea_merge_pull_request` takes no argument for a head. It reads the pull request's head at the moment of the merge, checks that head's statuses, and hands that same head to the forge. So it kept its own promise — only a checked head is merged — but nothing could tell it which head the review had approved, and a commit nobody reviewed landed.
## What changes
`gitea_merge_pull_request` takes **`head`**: the commit the caller reviewed, as a sha or at least its first 7 characters, compared case-insensitively. It is refused with nothing sent when `head` is missing, is not a commit, or is not the pull request's head; the refusal names the head the pull request has now. The forge is handed the full head as `head_commit_id`, so a push landing between this read and the merge is the forge's to refuse. The answer now carries `head`, and reads the pull request back after the merge: if the forge merged a different head anyway, the answer says so loudly (`landed`, `not_the_head_named`) rather than reporting the merge as the one asked for.
**`head` is required, not optional — a decision I made here, for the reviewer to judge.** An optional guard is the shape that let this happen. I found no code in any of the mesh's repositories (mesh-catalog, mesh-controller, mesh-tools, mesh-host, mesh-sdk, mesh-media-catalog) and no hq document that calls this tool; only agent sessions do, through the console. A session that leaves `head` out is told what to name.
## Not verified
- Whether Gitea 1.27.3 (the forge's version) actually refuses a merge whose `head_commit_id` differs from the branch's head at merge time. The tool already sent it before this change, and nothing here depends on it: the tool's own comparison is the guard; the forge's is the backstop for the read-to-merge window, and the read-back reports if it does not hold.
- **The tests were not run on this machine.** The module's local dependencies are stale (an older `@novox/mesh-sdk` without the `events`, `tools` and `provisioner` subpaths, and no `undici-types`), so it does not type-check here for reasons outside this change. `mesh/repo-check`'s TypeScript part, in the toolchain on the build node, is the first run of the new tests.
## Tests (test/guard.test.ts)
The fake forge's head is now a full 40-character sha a test can move, and the merge marks the pull request merged and can report a different head after it. Existing merge tests name the head they were reviewed at. New: no head, a blank head, a branch name and a 5-character prefix are refused; a head that moved after review is refused for both the full sha and its 7-character prefix, naming the new head, with nothing sent; a 7-character upper-case prefix merges and hands the forge the full sha; a forge that merges another head is reported as such.
mesh-controller#197 was reviewed at 065cfa0d, force-pushed to 3ced96f
at 12:14 UTC, checked green there at 12:22, and merged at 3ced96f at
12:26. The merge tool kept its promise that only a checked head is
merged, but it took no argument for the head the review approved, so it
merged whatever the head was and a commit nobody reviewed landed.
gitea_merge_pull_request now takes `head`, the commit the caller
reviewed (a sha, or at least its first 7 characters), and refuses with
nothing sent when it is missing, is not a commit, or is not the pull
request's head — naming the head it has now. The forge is handed the
same head as head_commit_id, so a push between the read and the merge
is the forge's to refuse. The answer reads back what was merged and says
so loudly if the forge merged another head anyway.
`head` is required, not optional: an optional guard is the shape that
let this happen. A caller that leaves it out is told what to name.
Deliverynovox/mesh-catalog@c87a4421b9c7 — delivered since 2026-10-11T00:02:20Z
Delivery plan — builds gitea → novox; no bus step
tier 0: gitea
novox: receives gitea
gitea provides npm-package-registry, git: its consumers are sent again after it
gitea keeps data (ADR 0233): what it holds is backed up before it moves
Machines
gitea on novox: passed — healthy 3 times over 2m30s
gitea on the rest running it: sent
Transitions
2026-10-10 23:36 (new) → proposed (announced): novox/mesh-catalog#193's head announced
2026-10-10 23:40 proposed → checked (checked): the verdict names this commit
2026-10-10 23:40 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail
2026-10-10 23:58 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move
2026-10-10 23:58 published → delivering (go): its walk started: let go by this owner in its group's order, by a person, or on the controller's own path — not a failed walk that carried a later commit for it, after which it waits for a walk of its own commit
2026-10-11 00:02 delivering → delivered (done): its walk is done, carrying its own commit: every machine of its deploy plan passed, or was left as its policy says
The commit's note under refs/notes/mesh-plan keeps every transition: git log --notes=mesh-plan.
<!-- mesh-delivery:view -->
**Delivery** `novox/mesh-catalog@c87a4421b9c7` — **delivered** since 2026-10-11T00:02:20Z
**Delivery plan** — builds gitea → novox; no bus step
- tier 0: gitea
- novox: receives gitea
- gitea provides npm-package-registry, git: its consumers are sent again after it
- gitea keeps data (ADR 0233): what it holds is backed up before it moves
**Machines**
- gitea on novox: passed — healthy 3 times over 2m30s
- gitea on the rest running it: sent
**Transitions**
- 2026-10-10 23:36 (new) → proposed (announced): novox/mesh-catalog#193's head announced
- 2026-10-10 23:40 proposed → checked (checked): the verdict names this commit
- 2026-10-10 23:40 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail
- 2026-10-10 23:58 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move
- 2026-10-10 23:58 published → delivering (go): its walk started: let go by this owner in its group's order, by a person, or on the controller's own path — not a failed walk that carried a later commit for it, after which it waits for a walk of its own commit
- 2026-10-11 00:02 delivering → delivered (done): its walk is done, carrying its own commit: every machine of its deploy plan passed, or was left as its policy says
The commit's note under `refs/notes/mesh-plan` keeps every transition: `git log --notes=mesh-plan`.
mesh/merge-gate: PASS — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea)
mesh/repo-check: PASS — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)
Delivery plan — builds gitea → novox; no bus step
tier 0: gitea
novox: receives gitea
gitea provides npm-package-registry, git: its consumers are sent again after it
gitea keeps data (ADR 0233): what it holds is backed up before it moves
Run by the build seat on shanks as build-1791675001130727927 (builds --log build-1791675001130727927).
--- the gate: gitea
gitea: ok, keeps data (valuable); WARNING: server stay(s) up and say(s) not how it is ready — judged by liveness alone, refused from 2026-11-18 (ADR 0240 rule 8)
long-running resources without health: 1
files asking for a setting without saying whether it is trusted: 0
1 manifest(s) checked. Judged against the seats this binary carries; a claim on one of the mesh's own seats is judged fully at registration, and a seat declared by a module not given here reads as unknown. Identities judged on a 6-character machine name, against the bounds of the providers given here
judging 1 module(s) changed against 4 machine(s), as the snapshot of 2026-10-10T23:27:48Z says they run
composing every machine as the mesh is
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
composing every machine with the change
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
merge gate: PASS — every machine composes with the change as it did without (4 of 4 compose)
judged by controller development build against the facts of 2026-10-10T23:27:48Z
modules the change moves:
- gitea (changed)
machines:
- a machine (f35): composes
- holds mesh-dns-resolver (idt): composes
- holds git, holds mesh-artifact-store, holds mesh-broker, holds mesh-catalog, holds mesh-delivery, holds mesh-dns-resolver, holds mesh-store, holds mesh-vault, holds npm-package-registry, the control node, the hub (ygcyr): composes
- a machine (yvbpsm): composes
a merge would rebuild 1 module(s) in 1 tier(s): gitea
--- its merge-check.sh (go toolchain)
ok git.novox.be/novox/mesh-catalog/checks/words 0.003s
words: 12 retired words and 70 daemons' lines, none in the 39388 strings of 654 files
ok git.novox.be/novox/mesh-catalog/checks/agenthome 0.012s
agent-home: no archive and no file written into under the agent's home
ok git.novox.be/novox/mesh-catalog/checks/paths 0.006s
paths: no module names another's state on its machine (3 known, listed to go)
long-running resources without health: 69, as kept
testing modules/gitea
ok gitea/cmd/npm-registry 1.782s
--- its merge-check-typescript.sh (typescript toolchain)
type-checking modules/gitea (TypeScript)
testing modules/gitea (TypeScript)
✔ a pull request opened after the margin is looked at, by its own update time (7.560155ms)
✔ every open pull request is read with what its base requires and what its head has (gitea_open_pulls) (41.357839ms)
✔ a pull request or issue number that is not one is refused before the forge is asked (1.375366ms)
✔ the newest status of each context is kept, with what it said, when and by whom (13.712596ms)
✔ a commit merges only when every required status is a success; a warning blocks (0.491415ms)
✔ no rule requires nothing; a rule that cannot be read says nothing, never yes (0.227662ms)
✔ a commit is contained in another when the compare finds nothing on it the other lacks (0.669918ms)
✔ a note line is appended once, as the forge's user, in its own repository (6.431953ms)
✔ a note that cannot be written is said, never read as written (0.404584ms)
✔ one view per pull request, found by its marker; only the mesh's statuses (0.525616ms)
✔ gitea_api refuses the forge's administration, a branch's protection and the merge check's status, and sends nothing (4.226677ms)
✔ gitea_api refuses a file written to the default branch or one a rule guards, and writes one on a branch of its own (43.901759ms)
✔ gitea_api still writes what an author writes (7.878229ms)
✔ gitea_branch_protection_set only strengthens (15.685798ms)
✔ gitea_merge_pull_request merges only a head whose required statuses all succeeded, and only that head (29.875799ms)
✔ gitea_merge_pull_request merges only the head it is named, and refuses one that moved after the review (12.923867ms)
✔ gitea_merge_pull_request says so when the forge merged a head other than the one it was handed (9.542658ms)
✔ gitea_delete_repo deletes only the tools' own user's repository, unguarded, untouched for 30 days, named in full (21.239701ms)
✔ gitea_api sends the path it judged, and refuses moving a pull request's base (2.486598ms)
✔ gitea_merge_pull_request merges only into a branch whose protection requires the checks (7.614797ms)
✔ a rule named for a branch is not made weaker than the glob rule that guards it now (13.497043ms)
✔ a repository owner or name is never another path (0.273884ms)
✔ the judges, without a forge (0.337144ms)
✔ gitea_api never acts as another user: sudo, in any case, and any query key it does not send are refused (2.66776ms)
✔ gitea_api never approves a pull request nor requests changes: those are the operator's (1.298344ms)
✔ no tool's caller text becomes another path or another user: `..` and `?sudo=` are refused before the forge (4.891885ms)
✔ every page of a pull request's files is read, though the forge caps a page at fifty (25.197616ms)
✔ a pass asks only the repositories that moved since the last look, every one before the first (0.362974ms)
✔ the directories above a merge's files that hold a module at the commit are said, the root never (11.438719ms)
✔ a lookup the forge refuses is thrown, not read as no module (3.720192ms)
✔ a renamed file is said under its old name too, as gone (novox/hq ADR 0267) (5.213509ms)
✔ the statuses asked for replace the rule's, and an empty list requires none (5.838686ms)
✔ a new rule refuses direct pushes unless asked; an administrator's override only when said (0.251423ms)
✔ each open pull request's new head is announced once, and nothing closed or merged (8.553437ms)
✔ a verdict is the head commit's status; an error is the forge's error, never a success (0.956711ms)
✔ each layer is its own status: the gate with the modules it judged, the repository's own check beside it (0.582967ms)
✔ no verdict sets warning on a required check; only a person's decision fails (0.8756ms)
✔ a commit status is set on the commit, under the merge check's context (33.924436ms)
✔ a change plan is the gate's result: said on the status and, when it builds something, as a comment (0.474876ms)
✔ first start: mints with the admin account, keeps the token at 0600, asks for the tools' scopes only (28.535274ms)
✔ second start: reuses the kept token, mints nothing (15.419555ms)
✔ the forge rejects the kept token (its data was restored): minted afresh, once, and the call goes through (17.60189ms)
✔ a token kept with write:admin by an earlier build is minted again without it at the next start, and the forge drops the old one (22.621547ms)
✔ a kept token lacking a scope the tools need: the forge refuses the call for the scope, the token is re-minted with the whole list, and the call goes through (18.609242ms)
✔ the kept file is gone but the forge still holds a token by that name: replaced, not refused (15.737069ms)
✔ concurrent first calls share one mint (10.055814ms)
✔ the admin account is refused: said plainly, nothing kept, and the next call fails the same way rather than crashing (12.417001ms)
✔ one process shares one source per kept file — the watcher and the tools never renew against each other (0.86332ms)
✔ a second process finds the token the first renewed, and reuses it instead of minting over it (17.984624ms)
✔ a configured token wins, and is reported rather than minted over when the forge rejects it (2.300536ms)
✔ nothing to mint with and no token: the client says what is missing (0.172402ms)
[gitea] no tools — no Gitea token, and nothing to mint one with — set MESH_GITEA_TOKEN, or MESH_GITEA_ADMIN_USER, MESH_GITEA_ADMIN_PASSWORD_FILE, MESH_GITEA_STATE_DIR
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-anUzqs/state/token
✔ the tools register once there is a way to a token, and the first call mints it (8.279824ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-lHf8ua/state/token
✔ a pull request can be closed, read and cleaned up through the tools (18.388069ms)
ℹ tests 53
ℹ suites 0
ℹ pass 53
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 334.141995
**Merge check** at `c32eb623`
- `mesh/merge-gate`: **PASS** — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea)
- `mesh/repo-check`: **PASS** — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)
**Delivery plan** — builds gitea → novox; no bus step
- tier 0: gitea
- novox: receives gitea
- gitea provides npm-package-registry, git: its consumers are sent again after it
- gitea keeps data (ADR 0233): what it holds is backed up before it moves
Run by the build seat on shanks as `build-1791675001130727927` (`builds --log build-1791675001130727927`).
```
--- the gate: gitea
gitea: ok, keeps data (valuable); WARNING: server stay(s) up and say(s) not how it is ready — judged by liveness alone, refused from 2026-11-18 (ADR 0240 rule 8)
long-running resources without health: 1
files asking for a setting without saying whether it is trusted: 0
1 manifest(s) checked. Judged against the seats this binary carries; a claim on one of the mesh's own seats is judged fully at registration, and a seat declared by a module not given here reads as unknown. Identities judged on a 6-character machine name, against the bounds of the providers given here
judging 1 module(s) changed against 4 machine(s), as the snapshot of 2026-10-10T23:27:48Z says they run
composing every machine as the mesh is
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
composing every machine with the change
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
merge gate: PASS — every machine composes with the change as it did without (4 of 4 compose)
judged by controller development build against the facts of 2026-10-10T23:27:48Z
modules the change moves:
- gitea (changed)
machines:
- a machine (f35): composes
- holds mesh-dns-resolver (idt): composes
- holds git, holds mesh-artifact-store, holds mesh-broker, holds mesh-catalog, holds mesh-delivery, holds mesh-dns-resolver, holds mesh-store, holds mesh-vault, holds npm-package-registry, the control node, the hub (ygcyr): composes
- a machine (yvbpsm): composes
a merge would rebuild 1 module(s) in 1 tier(s): gitea
--- its merge-check.sh (go toolchain)
ok git.novox.be/novox/mesh-catalog/checks/words 0.003s
words: 12 retired words and 70 daemons' lines, none in the 39388 strings of 654 files
ok git.novox.be/novox/mesh-catalog/checks/agenthome 0.012s
agent-home: no archive and no file written into under the agent's home
ok git.novox.be/novox/mesh-catalog/checks/paths 0.006s
paths: no module names another's state on its machine (3 known, listed to go)
long-running resources without health: 69, as kept
testing modules/gitea
ok gitea/cmd/npm-registry 1.782s
--- its merge-check-typescript.sh (typescript toolchain)
type-checking modules/gitea (TypeScript)
testing modules/gitea (TypeScript)
✔ a pull request opened after the margin is looked at, by its own update time (7.560155ms)
✔ every open pull request is read with what its base requires and what its head has (gitea_open_pulls) (41.357839ms)
✔ a pull request or issue number that is not one is refused before the forge is asked (1.375366ms)
✔ the newest status of each context is kept, with what it said, when and by whom (13.712596ms)
✔ a commit merges only when every required status is a success; a warning blocks (0.491415ms)
✔ no rule requires nothing; a rule that cannot be read says nothing, never yes (0.227662ms)
✔ a commit is contained in another when the compare finds nothing on it the other lacks (0.669918ms)
✔ a note line is appended once, as the forge's user, in its own repository (6.431953ms)
✔ a note that cannot be written is said, never read as written (0.404584ms)
✔ one view per pull request, found by its marker; only the mesh's statuses (0.525616ms)
✔ gitea_api refuses the forge's administration, a branch's protection and the merge check's status, and sends nothing (4.226677ms)
✔ gitea_api refuses a file written to the default branch or one a rule guards, and writes one on a branch of its own (43.901759ms)
✔ gitea_api still writes what an author writes (7.878229ms)
✔ gitea_branch_protection_set only strengthens (15.685798ms)
✔ gitea_merge_pull_request merges only a head whose required statuses all succeeded, and only that head (29.875799ms)
✔ gitea_merge_pull_request merges only the head it is named, and refuses one that moved after the review (12.923867ms)
✔ gitea_merge_pull_request says so when the forge merged a head other than the one it was handed (9.542658ms)
✔ gitea_delete_repo deletes only the tools' own user's repository, unguarded, untouched for 30 days, named in full (21.239701ms)
✔ gitea_api sends the path it judged, and refuses moving a pull request's base (2.486598ms)
✔ gitea_merge_pull_request merges only into a branch whose protection requires the checks (7.614797ms)
✔ a rule named for a branch is not made weaker than the glob rule that guards it now (13.497043ms)
✔ a repository owner or name is never another path (0.273884ms)
✔ the judges, without a forge (0.337144ms)
✔ gitea_api never acts as another user: sudo, in any case, and any query key it does not send are refused (2.66776ms)
✔ gitea_api never approves a pull request nor requests changes: those are the operator's (1.298344ms)
✔ no tool's caller text becomes another path or another user: `..` and `?sudo=` are refused before the forge (4.891885ms)
✔ every page of a pull request's files is read, though the forge caps a page at fifty (25.197616ms)
✔ a pass asks only the repositories that moved since the last look, every one before the first (0.362974ms)
✔ the directories above a merge's files that hold a module at the commit are said, the root never (11.438719ms)
✔ a lookup the forge refuses is thrown, not read as no module (3.720192ms)
✔ a renamed file is said under its old name too, as gone (novox/hq ADR 0267) (5.213509ms)
✔ the statuses asked for replace the rule's, and an empty list requires none (5.838686ms)
✔ a new rule refuses direct pushes unless asked; an administrator's override only when said (0.251423ms)
✔ each open pull request's new head is announced once, and nothing closed or merged (8.553437ms)
✔ a verdict is the head commit's status; an error is the forge's error, never a success (0.956711ms)
✔ each layer is its own status: the gate with the modules it judged, the repository's own check beside it (0.582967ms)
✔ no verdict sets warning on a required check; only a person's decision fails (0.8756ms)
✔ a commit status is set on the commit, under the merge check's context (33.924436ms)
✔ a change plan is the gate's result: said on the status and, when it builds something, as a comment (0.474876ms)
✔ first start: mints with the admin account, keeps the token at 0600, asks for the tools' scopes only (28.535274ms)
✔ second start: reuses the kept token, mints nothing (15.419555ms)
✔ the forge rejects the kept token (its data was restored): minted afresh, once, and the call goes through (17.60189ms)
✔ a token kept with write:admin by an earlier build is minted again without it at the next start, and the forge drops the old one (22.621547ms)
✔ a kept token lacking a scope the tools need: the forge refuses the call for the scope, the token is re-minted with the whole list, and the call goes through (18.609242ms)
✔ the kept file is gone but the forge still holds a token by that name: replaced, not refused (15.737069ms)
✔ concurrent first calls share one mint (10.055814ms)
✔ the admin account is refused: said plainly, nothing kept, and the next call fails the same way rather than crashing (12.417001ms)
✔ one process shares one source per kept file — the watcher and the tools never renew against each other (0.86332ms)
✔ a second process finds the token the first renewed, and reuses it instead of minting over it (17.984624ms)
✔ a configured token wins, and is reported rather than minted over when the forge rejects it (2.300536ms)
✔ nothing to mint with and no token: the client says what is missing (0.172402ms)
[gitea] no tools — no Gitea token, and nothing to mint one with — set MESH_GITEA_TOKEN, or MESH_GITEA_ADMIN_USER, MESH_GITEA_ADMIN_PASSWORD_FILE, MESH_GITEA_STATE_DIR
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-anUzqs/state/token
✔ the tools register once there is a way to a token, and the first call mints it (8.279824ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-lHf8ua/state/token
✔ a pull request can be closed, read and cleaned up through the tools (18.388069ms)
ℹ tests 53
ℹ suites 0
ℹ pass 53
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 334.141995
```
The review of #193 approved it with five nits, all taken.
Seven characters of a sha can be matched by a commit made to match them
in seconds, so an author who force-pushes could have a review merge a
commit nobody read. A head is now named by at least 12 characters, up to
64 for a sha-256 repository.
What landed was read from the forge's record of the pull request, which
is the very thing a race leaves behind. It is now read from the merge
commit: its second parent for a merge, the commit itself for a
fast-forward. A squash or a rebase makes new commits that name no head,
and the answer says it cannot be read back instead of staying silent.
A merge of another head than the one handed over was reported as a
field beside merged: true, which a caller reading only that would miss.
It is now an error that says the merge happened and what landed.
The refusal no longer asserts that the head moved after the review,
which it cannot know; it says the head is not the one named.
The second review found the gap it was asked to look for: everything
after the forge's merge could throw — a forge 5xx, a timeout, a commit
named without parents — and the caller got a bare error that never said
the merge had happened.
From the merge on, a read-back that fails or cannot be done is said
beside merged: true, in read_back; only a proven other head is an error,
and that error says the merge happened. A parent named without its sha
is a failed read-back, never an empty head said to have landed.
A fast-forward's commit is taken as the head only when the forge names
it so. That it does is assumed, not seen on this forge, so another sha
is said in read_back rather than raised as another head having landed.
The method's description lists all five methods and says which two can
be read back. The fast-forward test now gives the merge commit a
different second parent, so reading it as a merge would fail there.
mesh/merge-gate: PASS — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea)
mesh/repo-check: PASS — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)
Delivery plan — builds gitea → novox; no bus step
tier 0: gitea
novox: receives gitea
gitea provides npm-package-registry, git: its consumers are sent again after it
gitea keeps data (ADR 0233): what it holds is backed up before it moves
Run by the build seat on ace as build-1791675225821361448 (builds --log build-1791675225821361448).
--- the gate: gitea
gitea: ok, keeps data (valuable); WARNING: server stay(s) up and say(s) not how it is ready — judged by liveness alone, refused from 2026-11-18 (ADR 0240 rule 8)
long-running resources without health: 1
files asking for a setting without saying whether it is trusted: 0
1 manifest(s) checked. Judged against the seats this binary carries; a claim on one of the mesh's own seats is judged fully at registration, and a seat declared by a module not given here reads as unknown. Identities judged on a 6-character machine name, against the bounds of the providers given here
judging 1 module(s) changed against 4 machine(s), as the snapshot of 2026-10-10T23:35:44Z says they run
composing every machine as the mesh is
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
composing every machine with the change
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
merge gate: PASS — every machine composes with the change as it did without (4 of 4 compose)
judged by controller development build against the facts of 2026-10-10T23:35:44Z
modules the change moves:
- gitea (changed)
machines:
- a machine (f35): composes
- holds mesh-dns-resolver (idt): composes
- holds git, holds mesh-artifact-store, holds mesh-broker, holds mesh-catalog, holds mesh-delivery, holds mesh-dns-resolver, holds mesh-store, holds mesh-vault, holds npm-package-registry, the control node, the hub (ygcyr): composes
- a machine (yvbpsm): composes
a merge would rebuild 1 module(s) in 1 tier(s): gitea
--- its merge-check.sh (go toolchain)
ok git.novox.be/novox/mesh-catalog/checks/words 0.002s
words: 12 retired words and 70 daemons' lines, none in the 39396 strings of 654 files
ok git.novox.be/novox/mesh-catalog/checks/agenthome 0.011s
agent-home: no archive and no file written into under the agent's home
ok git.novox.be/novox/mesh-catalog/checks/paths 0.005s
paths: no module names another's state on its machine (3 known, listed to go)
long-running resources without health: 69, as kept
testing modules/gitea
ok gitea/cmd/npm-registry 1.721s
--- its merge-check-typescript.sh (typescript toolchain)
type-checking modules/gitea (TypeScript)
testing modules/gitea (TypeScript)
✔ a pull request opened after the margin is looked at, by its own update time (6.600909ms)
✔ every open pull request is read with what its base requires and what its head has (gitea_open_pulls) (42.001016ms)
✔ a pull request or issue number that is not one is refused before the forge is asked (1.415791ms)
✔ the newest status of each context is kept, with what it said, when and by whom (21.123539ms)
✔ a commit merges only when every required status is a success; a warning blocks (0.719513ms)
✔ no rule requires nothing; a rule that cannot be read says nothing, never yes (0.320112ms)
✔ a commit is contained in another when the compare finds nothing on it the other lacks (0.860948ms)
✔ a note line is appended once, as the forge's user, in its own repository (7.406683ms)
✔ a note that cannot be written is said, never read as written (0.449595ms)
✔ one view per pull request, found by its marker; only the mesh's statuses (0.697983ms)
✔ gitea_api refuses the forge's administration, a branch's protection and the merge check's status, and sends nothing (3.850327ms)
✔ gitea_api refuses a file written to the default branch or one a rule guards, and writes one on a branch of its own (39.304387ms)
✔ gitea_api still writes what an author writes (6.108101ms)
✔ gitea_branch_protection_set only strengthens (15.796385ms)
✔ gitea_merge_pull_request merges only a head whose required statuses all succeeded, and only that head (28.894918ms)
✔ gitea_merge_pull_request merges only the head it is named, and refuses one that is not the head (14.987093ms)
✔ gitea_merge_pull_request is loud when the forge merged a head other than the one it was handed (9.87834ms)
✔ gitea_merge_pull_request reads back a fast-forward, and says it cannot read back a squash or a rebase (37.142845ms)
✔ gitea_delete_repo deletes only the tools' own user's repository, unguarded, untouched for 30 days, named in full (23.334144ms)
✔ gitea_api sends the path it judged, and refuses moving a pull request's base (6.081162ms)
✔ gitea_merge_pull_request merges only into a branch whose protection requires the checks (6.718621ms)
✔ a rule named for a branch is not made weaker than the glob rule that guards it now (13.816502ms)
✔ a repository owner or name is never another path (0.235953ms)
✔ the judges, without a forge (0.332815ms)
✔ gitea_api never acts as another user: sudo, in any case, and any query key it does not send are refused (1.592414ms)
✔ gitea_api never approves a pull request nor requests changes: those are the operator's (2.392989ms)
✔ no tool's caller text becomes another path or another user: `..` and `?sudo=` are refused before the forge (5.050604ms)
✔ every page of a pull request's files is read, though the forge caps a page at fifty (27.990088ms)
✔ a pass asks only the repositories that moved since the last look, every one before the first (0.399692ms)
✔ the directories above a merge's files that hold a module at the commit are said, the root never (16.058377ms)
✔ a lookup the forge refuses is thrown, not read as no module (4.002745ms)
✔ a renamed file is said under its old name too, as gone (novox/hq ADR 0267) (5.48991ms)
✔ the statuses asked for replace the rule's, and an empty list requires none (4.981274ms)
✔ a new rule refuses direct pushes unless asked; an administrator's override only when said (0.258856ms)
✔ each open pull request's new head is announced once, and nothing closed or merged (8.38885ms)
✔ a verdict is the head commit's status; an error is the forge's error, never a success (0.837454ms)
✔ each layer is its own status: the gate with the modules it judged, the repository's own check beside it (0.671282ms)
✔ no verdict sets warning on a required check; only a person's decision fails (0.973801ms)
✔ a commit status is set on the commit, under the merge check's context (27.845455ms)
✔ a change plan is the gate's result: said on the status and, when it builds something, as a comment (0.433946ms)
✔ first start: mints with the admin account, keeps the token at 0600, asks for the tools' scopes only (27.084916ms)
✔ second start: reuses the kept token, mints nothing (14.940646ms)
✔ the forge rejects the kept token (its data was restored): minted afresh, once, and the call goes through (19.222236ms)
✔ a token kept with write:admin by an earlier build is minted again without it at the next start, and the forge drops the old one (21.237784ms)
✔ a kept token lacking a scope the tools need: the forge refuses the call for the scope, the token is re-minted with the whole list, and the call goes through (15.339374ms)
✔ the kept file is gone but the forge still holds a token by that name: replaced, not refused (16.244357ms)
✔ concurrent first calls share one mint (10.445996ms)
✔ the admin account is refused: said plainly, nothing kept, and the next call fails the same way rather than crashing (11.566984ms)
✔ one process shares one source per kept file — the watcher and the tools never renew against each other (0.777382ms)
✔ a second process finds the token the first renewed, and reuses it instead of minting over it (18.662344ms)
✔ a configured token wins, and is reported rather than minted over when the forge rejects it (2.476595ms)
✔ nothing to mint with and no token: the client says what is missing (0.198253ms)
[gitea] no tools — no Gitea token, and nothing to mint one with — set MESH_GITEA_TOKEN, or MESH_GITEA_ADMIN_USER, MESH_GITEA_ADMIN_PASSWORD_FILE, MESH_GITEA_STATE_DIR
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-XYN1ue/state/token
✔ the tools register once there is a way to a token, and the first call mints it (8.841751ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-MDD0rz/state/token
✔ a pull request can be closed, read and cleaned up through the tools (18.403396ms)
ℹ tests 54
ℹ suites 0
ℹ pass 54
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 336.682386
**Merge check** at `9ef810d9`
- `mesh/merge-gate`: **PASS** — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea)
- `mesh/repo-check`: **PASS** — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)
**Delivery plan** — builds gitea → novox; no bus step
- tier 0: gitea
- novox: receives gitea
- gitea provides npm-package-registry, git: its consumers are sent again after it
- gitea keeps data (ADR 0233): what it holds is backed up before it moves
Run by the build seat on ace as `build-1791675225821361448` (`builds --log build-1791675225821361448`).
```
--- the gate: gitea
gitea: ok, keeps data (valuable); WARNING: server stay(s) up and say(s) not how it is ready — judged by liveness alone, refused from 2026-11-18 (ADR 0240 rule 8)
long-running resources without health: 1
files asking for a setting without saying whether it is trusted: 0
1 manifest(s) checked. Judged against the seats this binary carries; a claim on one of the mesh's own seats is judged fully at registration, and a seat declared by a module not given here reads as unknown. Identities judged on a 6-character machine name, against the bounds of the providers given here
judging 1 module(s) changed against 4 machine(s), as the snapshot of 2026-10-10T23:35:44Z says they run
composing every machine as the mesh is
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
composing every machine with the change
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
merge gate: PASS — every machine composes with the change as it did without (4 of 4 compose)
judged by controller development build against the facts of 2026-10-10T23:35:44Z
modules the change moves:
- gitea (changed)
machines:
- a machine (f35): composes
- holds mesh-dns-resolver (idt): composes
- holds git, holds mesh-artifact-store, holds mesh-broker, holds mesh-catalog, holds mesh-delivery, holds mesh-dns-resolver, holds mesh-store, holds mesh-vault, holds npm-package-registry, the control node, the hub (ygcyr): composes
- a machine (yvbpsm): composes
a merge would rebuild 1 module(s) in 1 tier(s): gitea
--- its merge-check.sh (go toolchain)
ok git.novox.be/novox/mesh-catalog/checks/words 0.002s
words: 12 retired words and 70 daemons' lines, none in the 39396 strings of 654 files
ok git.novox.be/novox/mesh-catalog/checks/agenthome 0.011s
agent-home: no archive and no file written into under the agent's home
ok git.novox.be/novox/mesh-catalog/checks/paths 0.005s
paths: no module names another's state on its machine (3 known, listed to go)
long-running resources without health: 69, as kept
testing modules/gitea
ok gitea/cmd/npm-registry 1.721s
--- its merge-check-typescript.sh (typescript toolchain)
type-checking modules/gitea (TypeScript)
testing modules/gitea (TypeScript)
✔ a pull request opened after the margin is looked at, by its own update time (6.600909ms)
✔ every open pull request is read with what its base requires and what its head has (gitea_open_pulls) (42.001016ms)
✔ a pull request or issue number that is not one is refused before the forge is asked (1.415791ms)
✔ the newest status of each context is kept, with what it said, when and by whom (21.123539ms)
✔ a commit merges only when every required status is a success; a warning blocks (0.719513ms)
✔ no rule requires nothing; a rule that cannot be read says nothing, never yes (0.320112ms)
✔ a commit is contained in another when the compare finds nothing on it the other lacks (0.860948ms)
✔ a note line is appended once, as the forge's user, in its own repository (7.406683ms)
✔ a note that cannot be written is said, never read as written (0.449595ms)
✔ one view per pull request, found by its marker; only the mesh's statuses (0.697983ms)
✔ gitea_api refuses the forge's administration, a branch's protection and the merge check's status, and sends nothing (3.850327ms)
✔ gitea_api refuses a file written to the default branch or one a rule guards, and writes one on a branch of its own (39.304387ms)
✔ gitea_api still writes what an author writes (6.108101ms)
✔ gitea_branch_protection_set only strengthens (15.796385ms)
✔ gitea_merge_pull_request merges only a head whose required statuses all succeeded, and only that head (28.894918ms)
✔ gitea_merge_pull_request merges only the head it is named, and refuses one that is not the head (14.987093ms)
✔ gitea_merge_pull_request is loud when the forge merged a head other than the one it was handed (9.87834ms)
✔ gitea_merge_pull_request reads back a fast-forward, and says it cannot read back a squash or a rebase (37.142845ms)
✔ gitea_delete_repo deletes only the tools' own user's repository, unguarded, untouched for 30 days, named in full (23.334144ms)
✔ gitea_api sends the path it judged, and refuses moving a pull request's base (6.081162ms)
✔ gitea_merge_pull_request merges only into a branch whose protection requires the checks (6.718621ms)
✔ a rule named for a branch is not made weaker than the glob rule that guards it now (13.816502ms)
✔ a repository owner or name is never another path (0.235953ms)
✔ the judges, without a forge (0.332815ms)
✔ gitea_api never acts as another user: sudo, in any case, and any query key it does not send are refused (1.592414ms)
✔ gitea_api never approves a pull request nor requests changes: those are the operator's (2.392989ms)
✔ no tool's caller text becomes another path or another user: `..` and `?sudo=` are refused before the forge (5.050604ms)
✔ every page of a pull request's files is read, though the forge caps a page at fifty (27.990088ms)
✔ a pass asks only the repositories that moved since the last look, every one before the first (0.399692ms)
✔ the directories above a merge's files that hold a module at the commit are said, the root never (16.058377ms)
✔ a lookup the forge refuses is thrown, not read as no module (4.002745ms)
✔ a renamed file is said under its old name too, as gone (novox/hq ADR 0267) (5.48991ms)
✔ the statuses asked for replace the rule's, and an empty list requires none (4.981274ms)
✔ a new rule refuses direct pushes unless asked; an administrator's override only when said (0.258856ms)
✔ each open pull request's new head is announced once, and nothing closed or merged (8.38885ms)
✔ a verdict is the head commit's status; an error is the forge's error, never a success (0.837454ms)
✔ each layer is its own status: the gate with the modules it judged, the repository's own check beside it (0.671282ms)
✔ no verdict sets warning on a required check; only a person's decision fails (0.973801ms)
✔ a commit status is set on the commit, under the merge check's context (27.845455ms)
✔ a change plan is the gate's result: said on the status and, when it builds something, as a comment (0.433946ms)
✔ first start: mints with the admin account, keeps the token at 0600, asks for the tools' scopes only (27.084916ms)
✔ second start: reuses the kept token, mints nothing (14.940646ms)
✔ the forge rejects the kept token (its data was restored): minted afresh, once, and the call goes through (19.222236ms)
✔ a token kept with write:admin by an earlier build is minted again without it at the next start, and the forge drops the old one (21.237784ms)
✔ a kept token lacking a scope the tools need: the forge refuses the call for the scope, the token is re-minted with the whole list, and the call goes through (15.339374ms)
✔ the kept file is gone but the forge still holds a token by that name: replaced, not refused (16.244357ms)
✔ concurrent first calls share one mint (10.445996ms)
✔ the admin account is refused: said plainly, nothing kept, and the next call fails the same way rather than crashing (11.566984ms)
✔ one process shares one source per kept file — the watcher and the tools never renew against each other (0.777382ms)
✔ a second process finds the token the first renewed, and reuses it instead of minting over it (18.662344ms)
✔ a configured token wins, and is reported rather than minted over when the forge rejects it (2.476595ms)
✔ nothing to mint with and no token: the client says what is missing (0.198253ms)
[gitea] no tools — no Gitea token, and nothing to mint one with — set MESH_GITEA_TOKEN, or MESH_GITEA_ADMIN_USER, MESH_GITEA_ADMIN_PASSWORD_FILE, MESH_GITEA_STATE_DIR
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-XYN1ue/state/token
✔ the tools register once there is a way to a token, and the first call mints it (8.841751ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-MDD0rz/state/token
✔ a pull request can be closed, read and cleaned up through the tools (18.403396ms)
ℹ tests 54
ℹ suites 0
ℹ pass 54
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 336.682386
```
mesh/merge-gate: PASS — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea)
mesh/repo-check: PASS — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)
Delivery plan — builds gitea → novox; no bus step
tier 0: gitea
novox: receives gitea
gitea provides npm-package-registry, git: its consumers are sent again after it
gitea keeps data (ADR 0233): what it holds is backed up before it moves
Run by the build seat on novox as build-1791675368956133130 (builds --log build-1791675368956133130).
--- the gate: gitea
gitea: ok, keeps data (valuable); WARNING: server stay(s) up and say(s) not how it is ready — judged by liveness alone, refused from 2026-11-18 (ADR 0240 rule 8)
long-running resources without health: 1
files asking for a setting without saying whether it is trusted: 0
1 manifest(s) checked. Judged against the seats this binary carries; a claim on one of the mesh's own seats is judged fully at registration, and a seat declared by a module not given here reads as unknown. Identities judged on a 6-character machine name, against the bounds of the providers given here
judging 1 module(s) changed against 4 machine(s), as the snapshot of 2026-10-10T23:35:44Z says they run
composing every machine as the mesh is
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
composing every machine with the change
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
merge gate: PASS — every machine composes with the change as it did without (4 of 4 compose)
judged by controller development build against the facts of 2026-10-10T23:35:44Z
modules the change moves:
- gitea (changed)
machines:
- a machine (f35): composes
- holds mesh-dns-resolver (idt): composes
- holds git, holds mesh-artifact-store, holds mesh-broker, holds mesh-catalog, holds mesh-delivery, holds mesh-dns-resolver, holds mesh-store, holds mesh-vault, holds npm-package-registry, the control node, the hub (ygcyr): composes
- a machine (yvbpsm): composes
a merge would rebuild 1 module(s) in 1 tier(s): gitea
--- its merge-check.sh (go toolchain)
ok git.novox.be/novox/mesh-catalog/checks/words 0.003s
words: 12 retired words and 70 daemons' lines, none in the 39399 strings of 654 files
ok git.novox.be/novox/mesh-catalog/checks/agenthome 0.013s
agent-home: no archive and no file written into under the agent's home
ok git.novox.be/novox/mesh-catalog/checks/paths 0.006s
paths: no module names another's state on its machine (3 known, listed to go)
long-running resources without health: 69, as kept
testing modules/gitea
ok gitea/cmd/npm-registry 1.885s
--- its merge-check-typescript.sh (typescript toolchain)
type-checking modules/gitea (TypeScript)
testing modules/gitea (TypeScript)
✔ a pull request opened after the margin is looked at, by its own update time (13.260868ms)
✔ every open pull request is read with what its base requires and what its head has (gitea_open_pulls) (52.230913ms)
✔ a pull request or issue number that is not one is refused before the forge is asked (1.690252ms)
✔ the newest status of each context is kept, with what it said, when and by whom (20.632899ms)
✔ a commit merges only when every required status is a success; a warning blocks (0.628681ms)
✔ no rule requires nothing; a rule that cannot be read says nothing, never yes (0.27521ms)
✔ a commit is contained in another when the compare finds nothing on it the other lacks (0.865381ms)
✔ a note line is appended once, as the forge's user, in its own repository (8.554611ms)
✔ a note that cannot be written is said, never read as written (0.491761ms)
✔ one view per pull request, found by its marker; only the mesh's statuses (0.688271ms)
✔ gitea_api refuses the forge's administration, a branch's protection and the merge check's status, and sends nothing (4.305556ms)
✔ gitea_api refuses a file written to the default branch or one a rule guards, and writes one on a branch of its own (57.80564ms)
✔ gitea_api still writes what an author writes (8.952933ms)
✔ gitea_branch_protection_set only strengthens (17.119374ms)
✔ gitea_merge_pull_request merges only a head whose required statuses all succeeded, and only that head (34.895718ms)
✔ gitea_merge_pull_request merges only the head it is named, and refuses one that is not the head (17.584864ms)
✔ gitea_merge_pull_request is loud when the forge merged a head other than the one it was handed (10.809655ms)
✔ gitea_merge_pull_request says the merge happened when reading back what landed fails (26.422767ms)
✔ gitea_merge_pull_request reads back a fast-forward, and says it cannot read back a squash or a rebase (43.810931ms)
✔ gitea_delete_repo deletes only the tools' own user's repository, unguarded, untouched for 30 days, named in full (24.312454ms)
✔ gitea_api sends the path it judged, and refuses moving a pull request's base (4.350916ms)
✔ gitea_merge_pull_request merges only into a branch whose protection requires the checks (7.52939ms)
✔ a rule named for a branch is not made weaker than the glob rule that guards it now (13.460739ms)
✔ a repository owner or name is never another path (0.22736ms)
✔ the judges, without a forge (0.309471ms)
✔ gitea_api never acts as another user: sudo, in any case, and any query key it does not send are refused (3.741655ms)
✔ gitea_api never approves a pull request nor requests changes: those are the operator's (2.796794ms)
✔ no tool's caller text becomes another path or another user: `..` and `?sudo=` are refused before the forge (5.638288ms)
✔ every page of a pull request's files is read, though the forge caps a page at fifty (26.588096ms)
✔ a pass asks only the repositories that moved since the last look, every one before the first (0.451081ms)
✔ the directories above a merge's files that hold a module at the commit are said, the root never (15.978122ms)
✔ a lookup the forge refuses is thrown, not read as no module (5.407657ms)
✔ a renamed file is said under its old name too, as gone (novox/hq ADR 0267) (6.288258ms)
✔ the statuses asked for replace the rule's, and an empty list requires none (7.983971ms)
✔ a new rule refuses direct pushes unless asked; an administrator's override only when said (0.356161ms)
✔ each open pull request's new head is announced once, and nothing closed or merged (12.545648ms)
✔ a verdict is the head commit's status; an error is the forge's error, never a success (0.919842ms)
✔ each layer is its own status: the gate with the modules it judged, the repository's own check beside it (0.789411ms)
✔ no verdict sets warning on a required check; only a person's decision fails (1.055891ms)
✔ a commit status is set on the commit, under the merge check's context (37.470023ms)
✔ a change plan is the gate's result: said on the status and, when it builds something, as a comment (0.56909ms)
✔ first start: mints with the admin account, keeps the token at 0600, asks for the tools' scopes only (34.048457ms)
✔ second start: reuses the kept token, mints nothing (16.151722ms)
✔ the forge rejects the kept token (its data was restored): minted afresh, once, and the call goes through (62.593827ms)
✔ a token kept with write:admin by an earlier build is minted again without it at the next start, and the forge drops the old one (24.199574ms)
✔ a kept token lacking a scope the tools need: the forge refuses the call for the scope, the token is re-minted with the whole list, and the call goes through (83.281626ms)
✔ the kept file is gone but the forge still holds a token by that name: replaced, not refused (17.413884ms)
✔ concurrent first calls share one mint (10.576615ms)
✔ the admin account is refused: said plainly, nothing kept, and the next call fails the same way rather than crashing (12.987188ms)
✔ one process shares one source per kept file — the watcher and the tools never renew against each other (0.910071ms)
✔ a second process finds the token the first renewed, and reuses it instead of minting over it (18.122646ms)
✔ a configured token wins, and is reported rather than minted over when the forge rejects it (2.555243ms)
✔ nothing to mint with and no token: the client says what is missing (0.17222ms)
[gitea] no tools — no Gitea token, and nothing to mint one with — set MESH_GITEA_TOKEN, or MESH_GITEA_ADMIN_USER, MESH_GITEA_ADMIN_PASSWORD_FILE, MESH_GITEA_STATE_DIR
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-4H3K5O/state/token
✔ the tools register once there is a way to a token, and the first call mints it (8.555372ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-aTxbJe/state/token
✔ a pull request can be closed, read and cleaned up through the tools (20.278088ms)
ℹ tests 55
ℹ suites 0
ℹ pass 55
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 492.254154
**Merge check** at `c87a4421`
- `mesh/merge-gate`: **PASS** — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea)
- `mesh/repo-check`: **PASS** — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)
**Delivery plan** — builds gitea → novox; no bus step
- tier 0: gitea
- novox: receives gitea
- gitea provides npm-package-registry, git: its consumers are sent again after it
- gitea keeps data (ADR 0233): what it holds is backed up before it moves
Run by the build seat on novox as `build-1791675368956133130` (`builds --log build-1791675368956133130`).
```
--- the gate: gitea
gitea: ok, keeps data (valuable); WARNING: server stay(s) up and say(s) not how it is ready — judged by liveness alone, refused from 2026-11-18 (ADR 0240 rule 8)
long-running resources without health: 1
files asking for a setting without saying whether it is trusted: 0
1 manifest(s) checked. Judged against the seats this binary carries; a claim on one of the mesh's own seats is judged fully at registration, and a seat declared by a module not given here reads as unknown. Identities judged on a 6-character machine name, against the bounds of the providers given here
judging 1 module(s) changed against 4 machine(s), as the snapshot of 2026-10-10T23:35:44Z says they run
composing every machine as the mesh is
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
composing every machine with the change
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
the bus's user list leaves out 5 user(s) the mesh has minted no credential for: controller, node.f35, node.idt, node.ygcyr, node.yvbpsm. Each is a user that cannot connect until one is issued
merge gate: PASS — every machine composes with the change as it did without (4 of 4 compose)
judged by controller development build against the facts of 2026-10-10T23:35:44Z
modules the change moves:
- gitea (changed)
machines:
- a machine (f35): composes
- holds mesh-dns-resolver (idt): composes
- holds git, holds mesh-artifact-store, holds mesh-broker, holds mesh-catalog, holds mesh-delivery, holds mesh-dns-resolver, holds mesh-store, holds mesh-vault, holds npm-package-registry, the control node, the hub (ygcyr): composes
- a machine (yvbpsm): composes
a merge would rebuild 1 module(s) in 1 tier(s): gitea
--- its merge-check.sh (go toolchain)
ok git.novox.be/novox/mesh-catalog/checks/words 0.003s
words: 12 retired words and 70 daemons' lines, none in the 39399 strings of 654 files
ok git.novox.be/novox/mesh-catalog/checks/agenthome 0.013s
agent-home: no archive and no file written into under the agent's home
ok git.novox.be/novox/mesh-catalog/checks/paths 0.006s
paths: no module names another's state on its machine (3 known, listed to go)
long-running resources without health: 69, as kept
testing modules/gitea
ok gitea/cmd/npm-registry 1.885s
--- its merge-check-typescript.sh (typescript toolchain)
type-checking modules/gitea (TypeScript)
testing modules/gitea (TypeScript)
✔ a pull request opened after the margin is looked at, by its own update time (13.260868ms)
✔ every open pull request is read with what its base requires and what its head has (gitea_open_pulls) (52.230913ms)
✔ a pull request or issue number that is not one is refused before the forge is asked (1.690252ms)
✔ the newest status of each context is kept, with what it said, when and by whom (20.632899ms)
✔ a commit merges only when every required status is a success; a warning blocks (0.628681ms)
✔ no rule requires nothing; a rule that cannot be read says nothing, never yes (0.27521ms)
✔ a commit is contained in another when the compare finds nothing on it the other lacks (0.865381ms)
✔ a note line is appended once, as the forge's user, in its own repository (8.554611ms)
✔ a note that cannot be written is said, never read as written (0.491761ms)
✔ one view per pull request, found by its marker; only the mesh's statuses (0.688271ms)
✔ gitea_api refuses the forge's administration, a branch's protection and the merge check's status, and sends nothing (4.305556ms)
✔ gitea_api refuses a file written to the default branch or one a rule guards, and writes one on a branch of its own (57.80564ms)
✔ gitea_api still writes what an author writes (8.952933ms)
✔ gitea_branch_protection_set only strengthens (17.119374ms)
✔ gitea_merge_pull_request merges only a head whose required statuses all succeeded, and only that head (34.895718ms)
✔ gitea_merge_pull_request merges only the head it is named, and refuses one that is not the head (17.584864ms)
✔ gitea_merge_pull_request is loud when the forge merged a head other than the one it was handed (10.809655ms)
✔ gitea_merge_pull_request says the merge happened when reading back what landed fails (26.422767ms)
✔ gitea_merge_pull_request reads back a fast-forward, and says it cannot read back a squash or a rebase (43.810931ms)
✔ gitea_delete_repo deletes only the tools' own user's repository, unguarded, untouched for 30 days, named in full (24.312454ms)
✔ gitea_api sends the path it judged, and refuses moving a pull request's base (4.350916ms)
✔ gitea_merge_pull_request merges only into a branch whose protection requires the checks (7.52939ms)
✔ a rule named for a branch is not made weaker than the glob rule that guards it now (13.460739ms)
✔ a repository owner or name is never another path (0.22736ms)
✔ the judges, without a forge (0.309471ms)
✔ gitea_api never acts as another user: sudo, in any case, and any query key it does not send are refused (3.741655ms)
✔ gitea_api never approves a pull request nor requests changes: those are the operator's (2.796794ms)
✔ no tool's caller text becomes another path or another user: `..` and `?sudo=` are refused before the forge (5.638288ms)
✔ every page of a pull request's files is read, though the forge caps a page at fifty (26.588096ms)
✔ a pass asks only the repositories that moved since the last look, every one before the first (0.451081ms)
✔ the directories above a merge's files that hold a module at the commit are said, the root never (15.978122ms)
✔ a lookup the forge refuses is thrown, not read as no module (5.407657ms)
✔ a renamed file is said under its old name too, as gone (novox/hq ADR 0267) (6.288258ms)
✔ the statuses asked for replace the rule's, and an empty list requires none (7.983971ms)
✔ a new rule refuses direct pushes unless asked; an administrator's override only when said (0.356161ms)
✔ each open pull request's new head is announced once, and nothing closed or merged (12.545648ms)
✔ a verdict is the head commit's status; an error is the forge's error, never a success (0.919842ms)
✔ each layer is its own status: the gate with the modules it judged, the repository's own check beside it (0.789411ms)
✔ no verdict sets warning on a required check; only a person's decision fails (1.055891ms)
✔ a commit status is set on the commit, under the merge check's context (37.470023ms)
✔ a change plan is the gate's result: said on the status and, when it builds something, as a comment (0.56909ms)
✔ first start: mints with the admin account, keeps the token at 0600, asks for the tools' scopes only (34.048457ms)
✔ second start: reuses the kept token, mints nothing (16.151722ms)
✔ the forge rejects the kept token (its data was restored): minted afresh, once, and the call goes through (62.593827ms)
✔ a token kept with write:admin by an earlier build is minted again without it at the next start, and the forge drops the old one (24.199574ms)
✔ a kept token lacking a scope the tools need: the forge refuses the call for the scope, the token is re-minted with the whole list, and the call goes through (83.281626ms)
✔ the kept file is gone but the forge still holds a token by that name: replaced, not refused (17.413884ms)
✔ concurrent first calls share one mint (10.576615ms)
✔ the admin account is refused: said plainly, nothing kept, and the next call fails the same way rather than crashing (12.987188ms)
✔ one process shares one source per kept file — the watcher and the tools never renew against each other (0.910071ms)
✔ a second process finds the token the first renewed, and reuses it instead of minting over it (18.122646ms)
✔ a configured token wins, and is reported rather than minted over when the forge rejects it (2.555243ms)
✔ nothing to mint with and no token: the client says what is missing (0.17222ms)
[gitea] no tools — no Gitea token, and nothing to mint one with — set MESH_GITEA_TOKEN, or MESH_GITEA_ADMIN_USER, MESH_GITEA_ADMIN_PASSWORD_FILE, MESH_GITEA_STATE_DIR
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-4H3K5O/state/token
✔ the tools register once there is a way to a token, and the first call mints it (8.555372ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-admin" (write:repository, write:issue, read:user), kept at /tmp/gitea-aTxbJe/state/token
✔ a pull request can be closed, read and cleaned up through the tools (20.278088ms)
ℹ tests 55
ℹ suites 0
ℹ pass 55
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 492.254154
```
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Fixes issue 381 (a merge asked for a named head merged a newer head instead).
What happened, read from the forge
mesh-controller #197's timeline: the checks passed at 065cfa0d at 11:59 UTC (the head the review approved); the branch was force-pushed to 3ced96f at 12:14; the checks failed at an intermediate d0c9b9b4 at 12:17 and passed at 3ced96f at 12:22; the merge took 3ced96f at 12:26. Twelve minutes passed between the push and the merge, so this was not the forge's record of the head lagging the branch.
The report says the merge was asked for with head_commit_id 065cfa0d. It could not have been:
gitea_merge_pull_requesttakes no argument for a head. It reads the pull request's head at the moment of the merge, checks that head's statuses, and hands that same head to the forge. So it kept its own promise — only a checked head is merged — but nothing could tell it which head the review had approved, and a commit nobody reviewed landed.What changes
gitea_merge_pull_requesttakeshead: the commit the caller reviewed, as a sha or at least its first 7 characters, compared case-insensitively. It is refused with nothing sent whenheadis missing, is not a commit, or is not the pull request's head; the refusal names the head the pull request has now. The forge is handed the full head ashead_commit_id, so a push landing between this read and the merge is the forge's to refuse. The answer now carrieshead, and reads the pull request back after the merge: if the forge merged a different head anyway, the answer says so loudly (landed,not_the_head_named) rather than reporting the merge as the one asked for.headis required, not optional — a decision I made here, for the reviewer to judge. An optional guard is the shape that let this happen. I found no code in any of the mesh's repositories (mesh-catalog, mesh-controller, mesh-tools, mesh-host, mesh-sdk, mesh-media-catalog) and no hq document that calls this tool; only agent sessions do, through the console. A session that leavesheadout is told what to name.Not verified
head_commit_iddiffers from the branch's head at merge time. The tool already sent it before this change, and nothing here depends on it: the tool's own comparison is the guard; the forge's is the backstop for the read-to-merge window, and the read-back reports if it does not hold.@novox/mesh-sdkwithout theevents,toolsandprovisionersubpaths, and noundici-types), so it does not type-check here for reasons outside this change.mesh/repo-check's TypeScript part, in the toolchain on the build node, is the first run of the new tests.Tests (test/guard.test.ts)
The fake forge's head is now a full 40-character sha a test can move, and the merge marks the pull request merged and can report a different head after it. Existing merge tests name the head they were reviewed at. New: no head, a blank head, a branch name and a 5-character prefix are refused; a head that moved after review is refused for both the full sha and its 7-character prefix, naming the new head, with nothing sent; a 7-character upper-case prefix merges and hands the forge the full sha; a forge that merges another head is reported as such.
Delivery
novox/mesh-catalog@c87a4421b9c7— delivered since 2026-10-11T00:02:20ZDelivery plan — builds gitea → novox; no bus step
Machines
Transitions
The commit's note under
refs/notes/mesh-plankeeps every transition:git log --notes=mesh-plan.Merge check at
c32eb623mesh/merge-gate: PASS — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea)mesh/repo-check: PASS — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)Delivery plan — builds gitea → novox; no bus step
Run by the build seat on shanks as
build-1791675001130727927(builds --log build-1791675001130727927).Merge check at
9ef810d9mesh/merge-gate: PASS — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea)mesh/repo-check: PASS — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)Delivery plan — builds gitea → novox; no bus step
Run by the build seat on ace as
build-1791675225821361448(builds --log build-1791675225821361448).Merge check at
c87a4421mesh/merge-gate: PASS — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea)mesh/repo-check: PASS — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)Delivery plan — builds gitea → novox; no bus step
Run by the build seat on novox as
build-1791675368956133130(builds --log build-1791675368956133130).