docs(rules): Option B disclosure for shadow-star-shorthand citation in m-acc rule (r2; #4036 base-contamination fix)#4038
Merged
Conversation
…n m-acc rule (audit #4031 follow-up) Peer-Otto-Desktop's #4033 covered 3 of 5 audit findings + correctly identified one mis-attribution (my #4031 audit said the classifier-caught-otto memo was cited in holding-without-named-dependency rule; it's actually in persistence-choice-architecture rule — peer's "stale audit anchor" catch was right). The shadow-star-shorthand-autocomplete-marker.md case was correctly SKIPPED by peer's #4033 because that rule already has good user-scope disclosure at line 78. BUT the same memory file is also cited inline in `m-acc-multi-oracle-end-user-moral-invariants.md` line 191 — a second citation site my original audit's `sort -u` dedup hid. Peer-Otto-Desktop didn't pick this up because they followed my audit list directly. Fix: add Option B disclosure inline, pointing cold-boot agents at the in-repo shadow-star-shorthand rule that names the shorthand definitively. Audit-method gap revealed: future audits should track ALL rule→file edges, not unique file paths. Filed for future-Otto consideration. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
There was a problem hiding this comment.
Pull request overview
This PR cleanly re-lands a documentation/rule clarification for the m/acc rule, adding citation disclosure for a user-scope-only memory reference and pointing cold-boot agents to the in-repo shorthand rule.
Changes:
- Adds Option B disclosure for the
feedback_aaron_shadow_star_shorthand_*memory citation. - Points fresh-checkout readers to the in-repo
shadow-star-shorthand-autocomplete-marker.mdrule.
3 tasks
AceHack
added a commit
that referenced
this pull request
May 17, 2026
…bstrate surfaces (systemic, not rule-local) (#4041) Audit #4031 found 5 dangling rule→memory refs in .claude/rules/. Extending the audit to 4 other substrate surfaces revealed 29 total: 1 skill, 8 research, 17 backlog, 3 persona — 6× scaling. User-scope auto-load makes citations resolve transparently for Aaron's Otto-CLI; cold-boot agents on fresh checkouts (different machine, new contributor, CI agent) follow the citations and find nothing. Not P0 — system works for Aaron. But 29 vs 5 changes priority calculus. Filed as substrate-engineer candidate: `tools/hygiene/audit-dangling- memory-refs.ts` + CI integration would mechanize the discipline. Audit-method gap also documented: original audit's `sort -u` dedup hid multi-citation edges (caused #4036 to miss a citation site that #4038 r2 had to back-fill). Better audit form tracks file:line pairs, not deduplicated filenames. Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
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
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.
Summary
R2 re-land of #4036, which had branch-base contamination:
git diff origin/main..b/4036-HEADshowed 7 files differing because my branch was cut from an olderorigin/mainthan I thought, and peer-Otto-CLI's updates toB-0427/B-0476/B-0509/install.tslanded on main between fork and PR-open. From GitHub's perspective the PR's NET effect was: add m-acc edit + REVERT 3 unrelated peer changes. That's why Copilot's 5 threads flagged scope mismatch +last_updated+ broken-depends_on on the apparently-reverted files.Per
blocked-green-ci-investigate-threads.mdPattern "Re-land via cherry-pick", this PR is the clean re-land:origin/mainc56f7388→842abfa); 1 file changed, 4 insertions, 1 deletiongit diff origin/main..HEAD --name-onlyshows ONLY.claude/rules/m-acc-multi-oracle-end-user-moral-invariants.mdWhat the substrate change does
Adds Option B disclosure to
m-acc-multi-oracle-end-user-moral-invariants.mdline 191 — a second citation site offeedback_aaron_shadow_star_shorthand_*that my original audit (#4031)sort -udedup hid. This completes the audit→Option-B chain peer-Otto-Desktop's #4033 began.Test plan
git ls-remote origin memo/m-acc-shadow-star-disclosure-r2-... → 842abfa7...git diff origin/main..HEAD --name-onlyreturns only the m-acc rule file🤖 Generated with Claude Code