fix-data-repair
Repair data damaged by a bug — duplicates, wrong grants, orphans, unowned records, stale caches — with dry-run counts, backups, idempotent reversible scripts, and human approval for production. Use when root-cause analysis shows a bug may have written bad data.
75
94%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Passed
No findings from the security scan
Data Repair
Procedure
- Define the damage query and how you know a row is wrong.
- Dry run: counts and a sanitized sample (no personal data).
- Script: back up affected rows first, then repair. It must be idempotent (safe to rerun) and reversible.
- Staging: run it, compare counts before and after, rerun to prove idempotency.
- Production: request approval with the script, counts, and backup location. Never run it without a yes.
- Record customer impact (who got wrong access or charges) for an owner decision on communication.
- Verify by re-running the damage query: zero rows.
Exit criteria
Before/after counts, a backup reference, and a zero-row damage query, or a pending approval request.
Counterpart of tester skill: qa-data-integrity.
- Repository
itsablabla/qa-bug-agent-skills
- Revision
9a6f628
- Last updated
- First committed
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.