docs+ai-forge+merkle-dag: fix moved-spec links, complete CLAUDE.md layout, syntax-check all patterns, de-flake Case 13 - #197
Merged
Conversation
…DE.md layout - The four ai-forge phase design specs were moved from docs/specs/ to docs/history/specs/ in 0901c66 without updating their relative link to docs/ROADMAP.md. - CLAUDE.md's Layout omitted demo/ and workflows/, both CI-tested executable areas, and its Testing section claimed every package keeps tests under scripts/ (ai-native-memory uses tests/, demo uses test/). All package test suites, verify-contracts (313/313), the committed self-weave verification, and check-portable-paths pass unchanged. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01U142aAXSzf4pUt89HSvXTH
The check script ran node --check on patterns/rag.mjs and patterns/telos.mjs but not on the three later pattern modules, even though their tests were checked. Add each module ahead of its tests. ai-forge npm test passes; verify-contracts 313/313, the committed self-weave verification, and check-portable-paths are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01U142aAXSzf4pUt89HSvXTH
dsmcewan
changed the title
test-orchestrate.mjs Case 13 asserts the parallel run finishes before the serial one. With N=3 nodes and concurrency limit 2, the parallel run saves about one sleepMs; at 60 ms that saving was smaller than the run-to-run jitter of spawn and signing overhead, so the assertion flaked (CI: paraWall=540ms vs serialWall=515ms on node 22; locally 1 of 5 runs failed under full CPU load). Raise sleepMs to 400 so the saving dominates the jitter (10 of 10 loaded runs pass, smallest margin 271 ms). The assertion itself and the vPeak concurrency proof are unchanged; the header comment no longer claims an 80% threshold the code never enforced. merkle-dag and build-gate npm test, verify-contracts 313/313, and the committed self-weave verification pass. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01U142aAXSzf4pUt89HSvXTH
dsmcewan
marked this pull request as ready for review
September 24, 2026 15:42
Code reviewNo issues found. Checked for bugs and CLAUDE.md compliance. |
dsmcewan
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What changed
docs/history/specs/2026-06-29-ai-forge-phase-{a,b,c,c2}-design.md: theROADMAP.mdlink was../ROADMAP.md, left over from when these lived indocs/specs/(moved in 0901c66). Now../../ROADMAP.md. None of these files is hash-pinned.CLAUDE.md: the Layout list now includesdemo/andworkflows/; the Testing section no longer claims every package keeps tests underscripts/. No mythological names introduced.ai-forge/package.json: thecheckscript now runsnode --checkonpatterns/multiagent.mjs,eval.mjsandserving.mjs, not justrag.mjsandtelos.mjs.merkle-dag/scripts/test-orchestrate.mjs: Case 13's wall-clock floor (parallel run faster than serial) flaked in CI (paraWall=540msvsserialWall=515ms) because with 60 ms sleeps and limit 2 the parallel saving was smaller than spawn/signing jitter. Reproduced locally (1 of 5 runs under CPU load).sleepMsis now 400 so the saving dominates jitter: 10 of 10 loaded runs pass, smallest margin 271 ms. The assertion and the vPeak concurrency proof are unchanged.main(verify-stage work from build-gate: per-node verify stage, rebased onto main (supersedes #192) #195) to satisfy the up-to-date requirement.Verification
All 31 CI jobs and
required CIpass on 2d34b62; the Claude review found no issues. Locally: every packagenpm test,verify-contracts313/313, committed self-weave verification ok, portable paths 0 violations.Left for a maintainer
docs/clotho-phase-1-design.md:75cites a deleted file inside atext_sha256-pinned section; fixing it means regenerating the oracles.repository-manifest.jsonomits several components and is dated 2026-07-19 (CHANGE-PROTOCOL governance change).connectors/meta-ads-mcp/server.mjshas nopackage.json, tests, or CI job (F14, issue security(meta-ads-mcp): ungoverned runtime with a bypassable spend cap #177).workflows/hestia.jsuses a name absent fromdocs/mythological-vocabulary.md(F11); needs human approval.docs/STATUS.mdis dated 2026-07-19 and hash-pinned.🤖 Generated with Claude Code
https://claude.ai/code/session_01U142aAXSzf4pUt89HSvXTH