Codex review: version existence alone does not prove the npm registry
serves the tarball this tag's preflight built — the same version could
have been published from a different artifact and npm versions are
immutable. The resume resolver now downloads the preflight manifest,
requires its releaseSha to match the target, and compares the published
registry tarball's sha256 against the manifest before skipping the core
dispatch; mismatches abort with correction-tag guidance.
The 2026.7.1 retro measured ~13h tag-to-published for beta.2 and stable
2026.6.11; any post-npm failure previously hard-aborted retries because
the already-published guard refused the whole run.
- resolve_openclaw_npm_publish_state replaces the abort guard: an
already-published core version skips the core npm dispatch and resumes
the remaining stages; verify_published_release still proves the
registry state matches the tag before the page leaves draft.
- Windows and Android promotion runs concurrently with the core npm
publish (their only shared prerequisite is the draft release page,
which is now created before the dispatch) and each promotion
short-circuits when the release already carries its verified asset
contract, so retries only redo failed stages.
- The release proof cites the core npm run only when this run dispatched
one; resumed publishes rely on the registry package check.
Codex review + test findings on the pipeline convergence:
- candidate changelog provenance now validates the same section the
renderer publishes (correction tags may carry their own heading) via a
shared correctionVersionForTag helper
- guard_existing_public_release accepts a canonical proofless body with
intact dependency evidence, matching the renderer's documented
proof-omitted-at-limit state; retries re-append the proof
- ledgerFor regains main's noteReferences threading so prose-cited PRs
keep their contribution-record rows, and the ledger/verify tests pin
the merged entry shape (externalReferences) and highlight gate
Two competing release-notes pipelines existed: the release branch's
hardened render/verify/provenance pipeline (a486f3ab08 + dcee1da876,
battle-tested by 2026.7.1) and main's lighter prepare-github-release-notes
size gate (#103222). Repo policy is one canonical path; the release-branch
pipeline wins and main's unique value is grafted in:
- scripts/render-github-release-notes.mjs becomes the canonical release
body renderer (full/compact 125k char+byte modes, tag-pinned record
link, verification tail, canonical shipped-baseline format), now also
preferring a correction tag's dedicated changelog section (from
prepare's heading matrix).
- verify-release-notes.mjs is a three-way merge: release's --shipped-ref
cumulative baselines, provenance checks, highlights gate, and the
excluded-record rewrite fix, plus main's compact contribution rows,
externalReferences threading, and both-heading parser compat.
- release-candidate-checklist.mjs gains validateCandidateCheckout and
changelog-provenance gates that run before any dispatch.
- openclaw-release-publish.yml keeps main's fail-before-mutation early
notes gate (retargeted to the renderer) and adopts release's
render/verify_release_tag_target/canonical_release_body_matches flow.
- scripts/prepare-github-release-notes.mjs and its test are deleted;
release-notes-ledger.test.ts stays and pins the merged verify exports.
- .gitignore tracks every repo skill for Git-aware syncs; SKILL.md
runbooks and RELEASING.md document the converged contract.
Codex review findings: evidence reuse must not let a default-input run
stand in for a focused provider/mode/filter/package-spec run, and
recorded child runs can be re-run to failure while the parent stays
green. Manifests now record validationInputs; the resolver requires an
exact input match and healthy child runs, and the summary re-verifies
the chain root plus its recorded children. Reuse-mode evidence dispatch
notes now disclose the chain-root relationship.
* fix(opencode-go): remove deprecated mimo-v2-omni and mimo-v2-pro model aliases
These deprecated aliases reject agent requests from the OpenCode Go gateway.
Remove them from the provider catalog and clean up all references in probe
skip lists, CI workflows, and tests.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* fix(opencode-go): complete deprecated MiMo cleanup
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* fix(release): validate prepared npm package sets
Forward-port the beta3 package-set preflight, Telegram, and Parallels validation to main while preserving main's existing Bun install smoke. Recompute decoded proxy response lengths and require artifact tarballs to exactly match the verified manifest.
* fix(release): lock fixture registry proxy origin
* ci(mantis): add web ui chat proof lane
* ci(mantis): tighten web ui proof candidate parsing
* ci: tighten Mantis Web UI proof lane
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* refactor: extract reusable AI runtime package
* refactor: complete AI provider relocation
* refactor: keep llm core internal
* refactor(ai): make @openclaw/ai self-contained with host policy ports
Move pure transport helpers (tool projections, strict-schema normalization,
prompt-cache boundary, stream guards, anthropic/openai compat, request
activity) from src into packages/ai; move utf16-slice into
normalization-core. Inject host policy (guarded fetch, redaction,
strict-tool defaults, diagnostics logging) through AiTransportHost with
inert library defaults installed by src/llm/stream.ts. Narrow the public
barrel to instance-scoped createApiRegistry/createLlmRuntime; the
process-default runtime moves behind internal/ and
registerBuiltInApiProviders takes an explicit registry. Delete the
src/llm/api-registry re-export facade.
* fix(ai): teach node, jiti, and vite resolvers the @openclaw/ai and utf16-slice subpaths
The workspace alias tables in root-alias.cjs, plugin-sdk-native-resolver,
sdk-alias, the shared vitest config, and the Control UI vite config only
knew @openclaw/llm-core; Node-side plugin loading resolved @openclaw/ai
through the pnpm symlink to the unbuilt dist (checks-node-compact CI
failures), and the Control UI build broke on the new
normalization-core/utf16-slice subpath.
* chore(ui): drop leftover service-worker debug logging
* build(release): ship @openclaw/ai with its own shrinkwrap and honest dependency set
packages/ai declares only its six real runtime deps (kysely, chalk, json5,
tslog, zod, fs-safe, and proxyline were never imported); orphaned root deps
removed. generate-npm-shrinkwrap now treats publishable packages/* like
publishable plugins so the AI tarball pins its transitive tree even though
workspace deps are omitted from the root shrinkwrap. knip learns the
package entry points; the tsdown dts neverBundle option moves to its
documented deps.dts home; the README documents the no-semver internal/*
contract and host ports.
* docs(ai): add minimal external-consumer example app
examples/ai-chat consumes only the public @openclaw/ai surface (built dist
via the workspace link): isolated runtime, built-in provider registration,
one streamed completion. Supports Anthropic/OpenAI via env keys and a
keyless local Ollama target; live-verified against Ollama.
* docs(ai): document the @openclaw/ai package and workspace shrinkwrap boundary
* chore(check): include examples/ in duplicate-scan targets
* fix: emit normalization package subpaths
* fix: complete AI package boundary artifacts
* fix: align AI package boundary contracts
* fix(ci): stabilize package release contracts
* test: align documentation contract checks
* test: keep cron docs guard aligned
* test: align restored docs contract guards
* test: follow upstream docs contracts
* docs: drop superseded talk wording
Ensure Testbox wrapper workflows finalize backend sessions even when setup fails, align the check timeout fallback with the documented 120-minute default, and guard the workflow invariants.
