Conflicts, both sides kept: artifact-store-tools' invokes (the assigner's
and indexer's read tools, followed through views, beside the registry
seats' hold, release and remove), its main (the views reader and the
seat asker), its manifest row; gitea's invokes (the indexer's read tool,
the hand act and the builder's versions) and its consumes of the
indexer's views. gitea's account test takes the packages account's
names; repoCheckFacts, removed on main, is not imported. No catalogue
module reads the indexer's state.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
An image built unnumbered is published by its digest; the builder's
version.numbered now publishes it under its version, sending nothing.
The late publish moved into buildstep, so mesh-build-go and
mesh-build-image share one.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
mesh-sdk's publish package left mesh-sdk (ADR 0627): buildstep binds each
registry seat to its client in mesh-contracts, over mesh-sdk's
seat-neutral sender of parts. The late publish uses the same path.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
asked, waiting on the seat every five minutes, stored or refused, as the ask's
phase on its step key, so a publish or a copy that stands still is readable on
the bus. PublishOne and PublishFiles narrate for every build module; the image
copy and the late publish use the same narration.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
No build module writes into a store any more: mesh-build-go's store test
plays the holder with its own two-request upload, as mesh-build-node's does.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
Bundles go through oci-artifact-registry and answer the seat's reference;
a trunk package is packed under the builder's version and published through
npm-package-registry, so the holder can check it with the builder. Reads are
anonymous, so the write pair, the sealed secret and the store's address go.
The one-part sender moves into buildstep for both build modules, so the move
to the registry clients is made in one place.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
mesh-build-go consumes the builder's version.numbered and publishes the held
digest under its version, naming its one part by digest only, since the packed
bytes left the node; distribution tags it at the size its store holds, so the
version's manifest names the blob truly.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
The unit level fails on a file that differs, is missing or was not generated; a build
ask answers the same files in its out-of-step, a failed build's too, so a trunk build's
end raises generated-types-out-of-step (lead's ruling on A2k finding 1). The compare is
buildstep's, each build module its own root, with the seat generator as a requirement.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
Design 51 Publishing, ADR 0605 §3: a Go package is published from a trunk build through go-package-registry
under the version the builder gave it. The build packs the directory the artifact names as Go's module zip,
answers its digest and h1: sum, and on a version publishes it under the Go module path (the forge's Go seat
takes the name as the module path); a candidate publishes nothing. The ask's seat and version reach the build.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
ADR 0605 §9: a done key names the seat's reference, oci-artifact-registry/<module>/<artifact>@<digest>,
not the store's older artifact-store:// form; the builder keeps the older form beside it while a counted
node-engine reads only that (ADR 0624). The whole test reads it back as a node-engine routes it.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
Design 51: only a registry seat's holder writes into its store. The module pushed each bundle
and archive straight into the artifact store; it now publishes through the seat's publish and
publish-part with mesh-sdk's sender, invokes both verbs, and no longer requires artifact-store
or places its address. push.go is gone.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
ADR 0607 §3, §5: an asked semantic version (v0.1.15-feat-x.2) is downloaded through the Go seat and built whatever
go.mod names, said when go.mod names a newer one; one the seat does not hold fails naming it as a download that may be
retaken, never looked for among the pull request heads. Rows in the asked-version table test.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
With go-package-registry bound, the fetch's GOPROXY is the Go seat, then proxy.golang.org, then direct (kept for
this milestone, E11: removed by the first change after mesh-sdk's and mesh-contracts' first numbered versions
are in the Go seat); no GOPRIVATE, and the checksum database off for the forge's paths (design 51 *Go packages*,
ADR 0605 §6). The module requires and binds go-package-registry, so its new build waits for the forge's build
that offers it (E2). Unbound, the fetch is as before.
Claude-Session: https://claude.ai/code/session_01JG73KJNrH5EgFXMKKYSrig
mesh/delivery-group group feat/sdk1 rejected: composed together: the checker's verdict at the members' heads (5c08034be629): integration: meshbuildgo/cmd/mesh-…
mesh/check failed: integration: meshbuildgo/cmd/mesh-build-go… | 640455999e66 on ee75d4327ce5 | judged by the candidate's mesh-build-go at 785ebb388d5a
ensureModfile copied go.mod word for word into pins/<dir>/, where `go mod verify` runs, so a replace to ./x or ../x pointed
nowhere (route-proxy's failure of 2026-11-02 13:53, met a second time with the release's sibling packages). Every relative
replace target is now made absolute under the tree the go.mod came from. mesh-build-node installs a module's own npm
packages in a scratch copy of package.json, so a file:, link: or portal: dependency on a relative path would fail the same
way; it is refused naming the package and its path, as a module without a lock is.
Quick fix: rewrite the one replace route-proxy had, in ensureModfile (minutes). Thorough fix: the kind, a go.mod or
package.json copied away from its tree, for every replace form (single, block, with a version, ./ and ../) in the one place
all copies go through (ensureModfile, which versions.go and the pins share), and the same kind in the other build module
that copies a manifest. Chosen: thorough, since it was met twice and hides a failure until the delivery (quick-or-thorough,
fix-a-bug-met-twice). Test: TestARelativeReplaceStillResolvesInTheTemporaryModfile runs real go mod verify in the pins directory.
Claude-Session: https://claude.ai/code/session_01NcbgfJ9Gum47zkFKWRaovL
The test copies the node-engine's repository from beside the check (MESH_CHECK_BESIDE, mesh-host, with mesh-sdk's Go module
beside it as one workspace when a group places it), has the module build github.com/novox/mesh-host/cmd/mesh-compose there,
and runs it; it fails loudly when the repository is not beside.
Claude-Session: https://claude.ai/code/session_01NcbgfJ9Gum47zkFKWRaovL
The module already builds whatever repository, directory and package the ask names; the tests now say so for the node-engine's
repository (a Go module at its root, the command mesh-compose) instead of mesh-sdk's composer, which leaves mesh-sdk, and the
real-bus test builds a stand-in command in its fetched tree. The README tells the runner's new home and what a build says at
its end.
Claude-Session: https://claude.ai/code/session_01NcbgfJ9Gum47zkFKWRaovL
The digest counted a module outside the tree by its version and checksum, so every mesh-sdk merge invalidated each kept
verdict of each repository. A package of the mesh's forge is now hashed by the files the tests reach, as a replacement to a
directory already was; any other outside module still counts by version and checksum. Every kept verdict is invalidated
once, by the form change.
Claude-Session: https://claude.ai/code/session_01NcbgfJ9Gum47zkFKWRaovL
A build ask whose job built an artifact asks the new step build_go_imports once: `go list -deps` of each Go bundle's program,
no tests, the forge's packages outside the module's own directory (a sibling package of the same repository counts), sorted
and each once. The answer carries them as BuildAnswer.Imports. The step is also a tool, to read them by hand.
Claude-Session: https://claude.ai/code/session_01NcbgfJ9Gum47zkFKWRaovL
The runner, the toolchain guard, the fetch, the parsers, buildcpu, the conformance kit and mesh-pack move from mesh-sdk's
buildstep into modules/mesh-build-go/buildstep (own go.mod). mesh-build-go, mesh-build-node and mesh-build-image require it
by a same-repository replace; what the delivery parts read (the failure contract, the fetch request, the stale sweep, the
module tree shape, NewID, CapacitySetting) stays imported from mesh-sdk as sdkstep.
Claude-Session: https://claude.ai/code/session_01NcbgfJ9Gum47zkFKWRaovL
mesh/check checking fa6dc3a48b0c on 32c4c52261c4 | novox/mesh-catalog#594 in group feat/1126-removal
When the forge's listing of pull request heads fails, the fetch now says it
as a download failure that may be retaken, so a forge outage is asked again
and not held as a refusal (issue 1748, issue 1126 review 3).
Claude-Session: https://claude.ai/code/session_01NcbgfJ9Gum47zkFKWRaovL
The group's check has no name lookup, so go failed on the network before it
could say it knew no such revision, and the fetch, reading go's words, took
the asked head for a download failure. Whether a commit is a pull request's
head is now the forge's fact, asked whenever go cannot fetch it; go's words
only label a commit on no head (issue 1748, issue 1126 review 2).
Claude-Session: https://claude.ai/code/session_01NcbgfJ9Gum47zkFKWRaovL
A test that slept a fixed time and then asserted what a goroutine, process or
watcher did failed under the checker's load. Each now waits for its positive
evidence (a started baseline, a call counted, a state reached) within 30 s, and
the concurrency-limit tests meet at a barrier instead of overlapping by sleep.
Claude-Session: https://claude.ai/code/session_01NcbgfJ9Gum47zkFKWRaovL
ADR 0573 §2 and §4 have a build take the latest published version of
each mesh package, or a group's own head, whatever go.mod names (issue
1585, S15). The fetch writes each asked version into the directory's
temporary modfile before downloading: a trunk commit is required at its
Go version with its checksums for go mod verify, a pull request's head
is placed as a pin on a head is, a newer go.mod version is kept and
said, and the answer and the log name each.
Claude-Session: https://claude.ai/code/session_01NcbgfJ9Gum47zkFKWRaovL
The test built its module with no settings file and no Capacity; the new
buildstep.NodeCapacity answers 0 for that (no default names a node), so the
puller took no ask and the test timed out after 90 s in check 205f634c30df.
Failed first locally at origin/main af89893d (90.1 s), passes with the slot (2.5 s).
The levels a build module judges hand the tests MESH_BUILD_NODE=1; the tests that run a step in bubblewrap fail on it when bwrap is missing. Issue 1309
A replaced mesh-build-go left a check ask's step running and its start
sweep spared it (issue 1890, note 21). mesh-build-go, mesh-build-node and
mesh-build-image pin mesh-sdk 981a35b5, whose steps name their owner and
whose sweep ends another owner's at once; each runs the sweep through one
method that main calls and that its test runs through the conformance
kit's left-running case. The failurekit check now refuses a build module
whose tests do not run that case, test step or not.
go list leaves a file its build tags exclude (an integration test) in IgnoredGoFiles, which the digest did not hash;
mesh-provisioner-client's unit test globs *_test.go and reads two of them, so a change there reused a stale unit pass.
Review finding F2 on ba49a7f0.
47 packages carried the marker and were never kept: about 2000 tests per level ran on every check, whatever it
changed. The marker now lists, as paths or globs from the repository's root (? before a path whose absence is read),
what the package's tests read outside it; the digest hashes those files with the package's own non-Go files. A marker
that lists nothing, names an unclean path, or a line that matches nothing keeps today's never-kept digest.
The 46 markers list their reads; checks/deps and checks/trunk lose theirs (their tests read only temporary
directories). Digest form 6.
Quick fix of issue 1936, chosen by the operator 2026-10-26 21:56; the thorough fix (the Go test log's files kept with
each verdict) stays on the issue.
mesh/delivery-group group fix/capacity-follows-cpu-share rejected: composed together: the checker's verdict at the members' heads (63694fe9222d): unit: ..test
mesh/check checking 3f8b0b42480b on 7c5c29421f02 | novox/mesh-catalog#511 in group fix/capacity-follows-cpu-share
mesh-build-go, -node and -image declare build-hours (node=HH:MM-HH:MM in the node's local time,
empty = any hour) and hand it to the SDK puller (the operator, 2026-10-26: g14 builds only from
18:00 to 06:00); their SDK pin moves to mesh-sdk d00dac1, which has it.
Each of mesh-build-go, -node and -image declares cores-per-step (6, 2, 2) and hands it to the
SDK puller, which bounds the configured capacity by floor(slice quota cores / cores-per-step),
read live; thread limits and the status tools follow the effective capacity (issue 1806's
follow-up: three Go checks on shanks' 12-core slice starved past the 30-minute step bound).
mesh-build-go says its lint and test phases with the ask id (issue 1890).
mesh-build-node kept the first fifteen lines of node --test's output, so a passing line of the first file stood for the
failure. Both build modules now parse through the SDK (go test -json, node --test's TAP report written to a file, the
type-check and go vet as console failures), name themselves in the result, and run the SDK's captured cases; the
catalogue check failurekit refuses a node-build-* module with a test step whose tests do not run the kit.
jschoubben
mesh-tools
jochen