The bridge's whole test runs the program against an embedded bus as the node does: a start, a
merge, an admitted push to an unprotected branch and a restart after a move each give one
head-moved naming the head read through the seat git's branch and the head held before (start
true at a start); a head heard twice, a branch not watched and a restart at the head held give
none, and the bridge asks the seat git nothing but branch. Its bus permissions are composed for
the manifest as it now stands. The pull, open pull request, group and unprotected-branch tests
go; the owed test performs the gate's statuses once, after a restart and again by hand.
bridge_heads answers each key of the state heads with the head the bridge holds, read-only.
Builds: design 47 H §739 (how checked), §774; lead answers H-A-48, H-A-85 (1)
Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
B2: on every branch-moved of a watched branch, and at its start for every watched branch's head,
the bridge reads the head through the seat git's verb branch, as a kept call that waits on the
holder's return while the branch is watched, and writes one head-moved (repository, branch,
head, before the head held in its state heads, start) on its stream forge; a head equal to the
one held writes nothing. It reads no pull request, no tree, no file and no protection: the
merged and pushed messages, their inputs and the state inputs, the call of git.pull, the finding
branch-unprotected with its held finding writer, and the ten-second redelivery go. A move whose
head cannot be read is left unacknowledged for the bus to deliver again.
B3: the pull-message path (Convert, the open pull requests at start and reconnect, the
forge-delete message), the groups (internal/groups' group part, forge.group.merged, end-asks,
bridge_groups, bridge_group_end_again), bridge_merged_message with the forge-made reader, the
checker's owed and the merge act of a switched key (with the switch's bookkeeping read) go, with
their consumes, invokes, states and requires. The gate's owed path stays (H-A-71), its catch-up
from the gate's verb groups gone with that verb.
The two meet in main.go and the package compiles only with both, as it did not since B1. The
unit suites pass; the whole test (integration) is rewritten in B4.
Builds: design 47 H §739, §764 (the forge bridge; the branch protection); lead answers H-A-23, H-A-27, H-A-48, H-A-71
Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
The merged message and the catch-up asked the seat's groups with all=true, whose answer outgrew one bus message
(2888983 bytes, live 02:06), so no take-in answered. Read asks the open groups, MergedOf the one group its branch
names (group=<id>); the list a person asks (bridge_groups) keeps ReadAll.
Builds: lead: principle 7, live 02:06
Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
Its states' buckets, its stream's subjects and its finding observer take the module its credential names (the tools: the module the runtime serves them as) instead of constants of its name; a guard test finds none in its code.
Builds: design 47 G A-296, §428
Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
ForgeGroupEnd gains base; the bridge reads each member's base from the seat's
verb groups (Member.base is already in the contract and seatgen's output),
keeps groups.<group>.<base>, says one end per base from that base's members,
and bridge_group_end_again takes the base.
Builds: design 47 G §715, A-236
Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
The bridge watches the registrations' branches by origin (a first view of registrations, then registration-changed),
before the first registration the gate's source's (source-moved); no branch is named in its code, Trunk goes. Every
move is heard on the seat git's branch-moved through the bridge's durable on SEAT_git_branch-moved, its head read by
the seat's verb branch, never the event's: merged with its pull request and `base` (the watched branch) when the
seat's pull names the head a pull request's merge, else `pushed` with branch, head and `before`, both with the
take-in inputs at the head. The last head written per watched branch is kept in the state heads under
branch.<owner>.<name>.<escaped branch>, so a move heard twice writes once, and each watched branch's head is read at
the start and when a branch is registered. A move of an unprotected branch writes nothing. pull.merged into a watched
branch writes nothing (its message is the move's); the forge-made merge's trunk filter goes.
Builds: design 47 G §4, §7, §11, §20, §41, §507, §509, §532, §714, A-227, A-236
Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
The delivery repository's contracts and SDK live under modules/ as
mesh-delivery-contracts and mesh-delivery-sdk, their Go module paths
renamed to match the folders (the TypeScript names already did), and
every importer in mesh-delivery, mesh-catalog, mesh-host, mesh-tools
and mesh-sdk follows in the same group: a design as it should be, one
path per module, checked by the group's check building every importer.
Builds: design 47 G §427, §88
Claude-Session: https://claude.ai/code/session_01NcbgfGKaxw984XyAPfLm4h
Taken over from the join (6ada2089, f1417ab0, 6e162a60, 3a4e62b0, 73fb42d7, ee15a280, a5476f62, 94a4e62a, 33907a81, e519cc87, e8f268c1) and fitted to the final
contracts: the seat's events are pull-updated, pull-merged, pull-closed, forge-delete; OwedEnded keeps only the fields §164 names; merge-pull
takes no ask. A merge act is performed only as a switched act (§169); from G the gate owes none (§167).
Builds: design 47 G §43, §45, §157, §161, §162, §163, §164, §167, §168, §169, §171, §250
Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ
Taken over from the join's 473f9323, 5f40d230, 159a679b; the events are the seat's pull-updated, pull-merged, pull-closed and forge-delete.
Builds: design 47 G §252, §310, §312, §317
Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ
mesh/check failed (2): integration: mesh-delivery-assigner/cmd/mesh-delivery-assigner.TestAProposalIsAnswer…; and 1 more | af3b678a92f0 on 042b8d1fc160
holdLost builds a pending take-in naming neither WaitsFor nor Unknown, which TakeIn's check refused, so the hold stayed
in the journal and the watchman never saw it.
Quick fix: set Unknown on holdLost's event. Thorough fix (chosen): the check accepts a pending revision that names Lost,
the field made for it; one naming nothing is still refused. Tests: a row accepting Lost, the refusal row kept, and the
lost-hold scenario encodes its said event.
Claude-Session: https://claude.ai/code/session_01NcbgfFSnn24M3GcqttoU7i
Both ForgeMade and Tags took the tool's wrapper for the pull request / the tag list, so every forge-made merge read
"#0 is not merged" and bridge_tags failed, leaving builds unnumbered.
Quick fix: unwrap "result" in forgeAPI (the one Go reader of gitea_api).
Thorough fix (chosen, the kind was met twice and hid behind a drifted fake): bridge.API answers what the tool answers,
forgeBody is the one reader of that shape and refuses an answer without a result, naming it; ForgeMade refuses number
0 as "no pull request"; the unit fakes answer the tool's real wrapped shape, with rows for the bare body.
Claude-Session: https://claude.ai/code/session_01NcbgfFSnn24M3GcqttoU7i
mesh/check failed (2): mesh-build-go: the indexer's judge at the group's mesh-delivery head 7b7096f37ea76e1…; and 1 more | 7b7096f37ea7 on ae31b66a9e33
- A take-in releases its claims (compare-and-set, only while the claim is still its own) on a lost claim, on a claim it
cannot read back, and on a merge it fails to take in: a claim left behind refused every later take-in until a
restart. T5 row with a store that fails once, both ways.
- The start step waits for nothing: with the build it runs unknown it says so and the forge consumer starts; a merged
message left unacknowledged asks the build again (clause 1). T10 row: the start answers at once and takes nothing in.
- A toolchain report evaluates pending again (ToolchainReported), since the build a revision names moves with the
toolchain. T7 row: the revision is taken in once the report makes it name the running build.
- The group's end is told by its whole subject (process.ModuleStreamSubject), so a repository named forge/group is
never taken for one; row in the indexer's handle test.
- The watchman asks the indexer's waits through indexer_pending at its start (invokes), before its durable consumer,
which does not hear again what it acknowledged before a restart; the comment that said the whole stream is heard is
corrected. W2 row: a wait seeded within its bound is raised past it.
- bridge_groups names each group's last end ask, its who, why and when, and the bridge's answer (state end-asks).
- The check's words for the indexer's rules: the trunk's manifest validator.
- The bridge's end drops the merged-state test its merge commit already implies.
- Test budget: groupask_test folded into groups_test, and the K1 mesh/check test into the owed status one: 804 unit
tests (was 806).
Claude-Session: https://claude.ai/code/session_01JG73KG7AfSdh6FP8VrBqCn
The bridge consumes the forge module's forge.delete and writes a branch-deleted message on FORGE for a deleted
branch (never the trunk, never a tag), so the builder removes the branch's tagged versions (design 51 *The forge
bridge says a branch's end*). Only the forge module's account may say forge.delete: the bus refuses it from
another (a case of the bridge's whole-module test). bridge_tags, served by a new tools bundle, answers a
package's git tags by prefix through the forge module's gitea_api, for the builder's floor.
One unit test merged into another, so the tags test adds no function (test budget).
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig