Repo Index Drift Check
Skill1 starsChanged 4 months agoMITView on GitHub
Audit a hub/index repo's claims (skill counts, descriptions, links) against what the linked repos actually contain right now, and flag any drift. Use periodically on any README that summarizes other repos' contents, especially after any linked repo changes, don't trust the index's own stated numbers as ground truth.
In this repository25 other files it gives its agents.
- templates/AGENTS.mdAGENTS.md
- braindump-autoSkill
- braindumpSkill
- Capability Claim GroundingSkill
- Derived Figure AuditSkill
- design-review-loopSkill
- Dictation Garble CatcherSkill
- Discord Todo OpsSkill
- Full Account Security AuditSkill
- Large Task Session SplitSkill
- Personal Dashboard StyleSkill
- Pre-Push Secret ScanSkill
- Project Design DocSkill
- Public Repo Leak RetractionSkill
- repo-hygieneSkill
- Safe Section DeletionSkill
- Session HandoffSkill
- Skill Overlap AuditSkill
- superbraindumpSkill
- ui-change-visual-verifySkill
- Verify, Don't TrustSkill
- Windows Process Restart (Safe)Skill
- Windows Shell Tool SelectionSkill
- Worktree Task-Pack VerificationSkill
- Zombie Process SweepSkill
The file
Files
Repo Index Drift Check
An index repo's claims about the things it links to (a count, a one-line description) are a snapshot from whenever they were last written, and they go stale the moment the linked repo changes without a corresponding update here.
Procedure
1. Parse the index repo's claims about each linked repo: stated counts, descriptions, feature callouts.
2. Independently verify the real current state of each linked repo, never trust the index's number as the source of truth:
gh api repos/{owner}/{repo}/git/trees/main?recursive=1 --jq '.tree[] | select(.path | test("SKILL.md$"))' | wc -l
gh repo view {owner}/{repo} --json description3. Diff claimed vs. actual. Any mismatch is drift: report it plainly (what the index says vs. what's actually there).
4. Fix mismatches at the source, not just the index: a stale count is a symptom; check whether the underlying repo's own README/description also drifted from its own real content (it's common for both the repo's own claims AND the index pointing to it to be wrong simultaneously).
5. Re-run after any change to a linked repo, not just once. This is a check to repeat, not a one-time fix.
Real precedent
In one afternoon, the actual number went 285 (what the index and both repos claimed) → 289 (the real, verified count) → 287 (after trimming two genuinely useless skills). Three different true numbers in a few hours. A repo's own stated count is a claim, not a fact, and it drifts faster than it looks like it would.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

