Skip to content

memories, fully-clean.cases: cross-link the two superseded-head records - #2916

Merged
d-morrison merged 1 commit into
mainfrom
claude/cross-link-superseded-head
Sep 1, 2026
Merged

memories, fully-clean.cases: cross-link the two superseded-head records#2916
d-morrison merged 1 commit into
mainfrom
claude/cross-link-superseded-head

Conversation

@d-morrison

Copy link
Copy Markdown
Collaborator

Closes #2912.

What

#2907 added a bullet to memories/github-mcp-tools.md on check_suite.completed events that name a superseded head, written as if the corpus had no prior record. shared/workflow/fully-clean.cases.md already carried one — ## A check_suite.completed wake at a superseded head (ucdavis/bcs#732, 2026-08-23). Neither referenced the other.

This adds one pointer in each direction, as real relative links so check-links.py validates them, and deletes nothing.

Why not consolidate

The two records measure different sessions and repositories, and the memory bullet's substantive content — the corrected "webhooks do not deliver CI success" claim, the App-suite and cancelled-suite exclusions the event declares, the lower-bound framing, and the cross-repo occurrences — is not in the case. challenge-redundant-content says merge only when nothing is lost; here something would be.

How the gap happened

The absence was asserted from a grep of memories/ plus a targeted read of one bullet — a phrase search over a subset returned nothing and was read as "the corpus lacks this". That is the grep-is-not-coverage failure verbatim, and it was mine, on #2907.

Verification

new-line-breaks run after committing, with a negative control (appended two-sentence line → exit 1; reverted → exit 0). No non-ASCII in added lines. check-links.py: no broken relative links, which covers both new cross-file links.

Disjoint from open #2913 (preferences.md, ardia, gia) — no merge-order constraint.

🤖 Generated with Claude Code

https://claude.ai/code/session_017trXKZFoCYC4kP7NKNnN6F


Generated by Claude Code

Closes #2912.

#2907 added a bullet to memories/github-mcp-tools.md on check_suite.completed
events that name a superseded head, written as if the corpus had no prior
record. fully-clean.cases.md already carried one (ucdavis/bcs#732,
2026-08-23). Neither referenced the other. The gap was asserted from a grep
of memories/ and a targeted read, which is the grep-is-not-coverage failure.

Add one pointer in each direction and delete nothing: the two records
measure different sessions and repositories, and the memory bullet's new
content (the corrected "no CI success events" claim, the App-suite
exclusions, the cross-repo occurrences) is not in the case.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017trXKZFoCYC4kP7NKNnN6F
@github-actions

github-actions Bot commented Sep 1, 2026

Copy link
Copy Markdown
Contributor

Claude finished reviewView run


No prior comments on this PR. This is a small, self-contained documentation change adding a bidirectional cross-reference between two existing case records about check_suite.completed events naming a superseded head — one in memories/github-mcp-tools.md and one in shared/workflow/fully-clean.cases.md.

I verified both quoted section titles ("A check_suite.completed wake at a superseded head" and "A check_suite.completed event can name a superseded head") exist verbatim in their respective files, and both relative markdown links (../shared/workflow/fully-clean.cases.md from memories/, ../../memories/github-mcp-tools.md from shared/workflow/) resolve correctly to real files at the expected repo-root-relative paths.

This is exactly the kind of change the repo's own run-ums-proactively.md / dupe-check discipline calls for: rather than duplicating the incident narrative in both places, it links the two so a reader following either lands on the other, and it's honest about the fact that the second bullet was written without having found the first ("this bullet was written without having found; read the two together"), consistent with learn-from-review-findings norms in this corpus.

No factual, structural, or style problems found:

  • Cross-reference targets and quoted titles both check out against current file contents.
  • No code touched, so no CLAUDE.md coding-style/practices rules apply.
  • No prose ambiguity, broken markup, or line-wrap issues in the added lines.
  • PR is open, not a draft, mergeable, and has no prior review comments to reconcile against.

Verdict

Ready for merge

Structured Review Data (JSON)
{
  "schema_version": "1.0",
  "reviewer": "claude",
  "commit_sha": "121d99a",
  "verdict": "CLEAN",
  "findings": []
}

Reviewed commit: 82c82b0

@github-actions

github-actions Bot commented Sep 1, 2026

Copy link
Copy Markdown
Contributor

💰 Cost: $0.5194 (review) — run

@d-morrison
d-morrison merged commit 99f18d4 into main Sep 1, 2026
14 checks passed
@d-morrison
d-morrison deleted the claude/cross-link-superseded-head branch September 1, 2026 20:37
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Cross-link the #2907 superseded-head memory with the existing fully-clean.cases.md case

2 participants