chore(memory): promote strong-fit configuration drafts from memory-pipeline for review#130
chore(memory): promote strong-fit configuration drafts from memory-pipeline for review#130Hareet wants to merge 15 commits into
Conversation
LLM_PROVIDER=claude-cli runs filter/distill via 'claude -p' on the operator's subscription (no API key). isRateLimitError/isAuthError/isBatchFatalError detect global LLM failures so the runner can stop the batch instead of flagging each PR. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…auth --concurrency (PIPELINE_CONCURRENCY, [1,10]) worker pool; --last removes the ~100-PR per-run cap; --resume skips processed PRs; --force bypasses filter. Rate-limit/auth errors stop the batch (exit 2) with a resume hint. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
… gate; pin Node 22 CI validates the domains/ corpus on PRs and fresh drafts in the pipeline run. .env.example set for claude-cli + Opus 4.8/max + concurrency 4; .nvmrc=22.18.0. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…ructure to VALID_DOMAINS Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…ft with a test Derive CHTDomain/CHTService/CHTWorkflow from the CHT_DOMAINS/CHT_SERVICES/CHT_WORKFLOWS const arrays; pipeline.ts re-exports instead of re-declaring; distiller imports CHT_SERVICES. New test asserts schema.json enums == the const arrays. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
schema.json enums, doc-search Record maps, TEMPLATE.md, and distiller/domain-inference tests. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…n't rejected Both VALID_DOMAINS hand-lists now derive from CHT_DOMAINS; updates the validator test message. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…ons, re-export) Shared linear JSON extractor; CC refactors + justified NOSONAR; more Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…R not honored) Extracted helpers; SonarCloud Automatic Analysis ignores the NOSONAR block, so refactored to <=5. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…rameters) logPrStart/processBatchItem/runWorker now take a shared BatchCtx, each <=4 params; CC still <=5. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…rage exclude Proxyquire tests for the run-pipeline helpers + concurrency>1; export 3 helpers for testing. Only providers/open-review-pr/domain-inference stay excluded; coverage gate green Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Hareet
force-pushed
the
seeding-claude-cli-v2
branch
from
1979c05 to
6a2742a
Compare
sugat009
left a comment
sugat009
left a comment
There was a problem hiding this comment.
Requesting changes, though to be clear the fix is upstream and not something to patch by hand here. This seeder faithfully promotes what the distillation pipeline generated, so nothing is wrong with the promotion itself. The problem is in the drafts the pipeline produced, and I verified it against the live medic/cht-core API for every draft in this PR.
The defect: issueNumber / issueUrl name the merge PR, not the resolved issue
One precise note up front, because the intuitive version of this is not what is happening: issueNumber and issueUrl always agree with each other (the URL is literally /issues/<issueNumber>). The bug is that in 4 of 10 drafts here, that shared number is a pull request, so /issues/N silently redirects to /pull/N. The real resolved issue survives only in the PR-title slug. For example:
10198-fix8026-validate-empty-branding-doc.mdstoresissueNumber: 10198/issueUrl: .../issues/10198, but 10198 is the merge PR ("fix(#8026): validate empty branding doc"). The real issue is #8026 (closed): "Admin app fails to updatebrandingif existing document is malformed".9407-fix9406-disable-unsupported-languages.mdstoresissueNumber: 9407/issueUrl: .../issues/9407, but 9407 is the merge PR ("fix(#9406): disable unsupported languages"). The real issue is #9406 (closed): "Update default configurations to disable unsupported languages".
This is the same scraper behaviour flagged on #121, and it is pipeline-wide: across the four clean seeders, 60 of 107 drafts are affected.
Why request-changes rather than merge-and-fix-later
This corpus is the agent's memory of resolved issues (consumed by the Context Analysis Agent, see #135). A reference that resolves to a PR instead of the issue is simply wrong data, and it specifically defeats #135's planned de-duplication by issue id, since the "id" ends up being the PR id. Regenerating the drafts after the upstream fix will rewrite most of these files anyway, so merging now would churn the corpus twice.
Suggested fix (at the source, not file by file)
- Fix the distiller/scraper to take the resolved issue from the
type(#N):PR title (the filename slug already extracts it correctly), and keep the PR number insource_prwhere it belongs. - Regenerate and re-promote this domain's drafts.
- Add de-duplication by real issue id (also a #135 acceptance item).
Open to discussing the approach. Once the drafts carry the real issue references, this should be a quick re-review.
2 participants
Promotes 10 strong-fit
configurationdrafts fromagent-memory/_pending/intoagent-memory/domains/configuration/issues/for squad content review.Categories: feature (6), bug (4)
Themes: translations & RTL support, branding/partners doc validation, header-tab ordering, UI-extension service-worker rebuilds.
All 10 carry
domainFit: strong+ a## Domain Rationalesection. 18 weak-fit drafts deferred to Stream C —configurationwas a catch-all in this run (64% weak), so this PR is intentionally the distilled minority.seeding-claude-cli-v2(feat(#108): seeding pipeline - CLI provider, domain-rationale, infrastructure domain, concurrency #119) — retarget tomainafter the schema lands.validate-schema: passing, 0 failures.