app-audit
Skill0 starsChanged 5 months agoLicence not identifiedView on GitHub
Run a rigorous, repeatable, convergent audit of a codebase covering schema integrity, data flow, security, concurrency, resource bounds, spec compliance, operational readiness, test coverage with spec→acceptance-test mapping, and diagnosability. Three profiles — "quick audit"/"sanity check" (detector-grounded minutes-long pass, no setup), plain "audit" (standard), "deep audit"/"release audit" (everything) — so the ceremony matches the moment. Use whenever the user asks to audit, review, QA, verify, or validate a codebase — including reviewing a specific PR, branch, or diff (scopes to the change-set, can post findings as PR comments / GitHub issues) — especially before a release or after a major refactor. Consumes cartographer's codemap for targeted retrieval; produces a persistent AUDIT_LOG.md so audits converge across rounds. Self-learning: recurring findings get codified into checklist items and learned detector rules, a per-repo chronicle carries lessons and cleared findings forward between rounds, and project-agnostic lessons can distill (privacy-scrubbed, opt-in) to a shared brain. Interrupted rounds resume from .audit/run_state.json ("resume the audit"); scope ranks files by churn × coverage hotspots; each round refreshes TECHNICAL_DEBT.md, machine-readable .audit/status.json (with per-round timeline for CI), and an optional self-contained HTML dashboard. Opt-in multi-model layer: cross-provider blind challenges and second opinions via codex/gemini/ollama CLIs when present (explicit consent before code leaves the machine; Ollama stays local), plus a consensus gate for "multi-model review" requests. Spec surface: diff-driven claim-level spec-drift detection (code and spec disagreeing is a finding either way) and a per-round traceability matrix (spec § ↔ code ↔ test ↔ findings). Self-evaluating: a benchmark harness (benchmark/) scores the audit's own recall/precision/false-positive rate against seeded-bug fixtures, with a regression gate for CI. Detector-grounded: a readiness probe (scripts/greenloop-doctor.sh) reports which analyzer tiers are live so a missing one degrades findings visibly, not silently, and a bundled offline semgrep ruleset gives SAST grounding on a bare machine. CI-native: findings export to SARIF 2.1.0 (GitHub code-scanning / IDE Problems panel) and GitLab Code Quality, with ready-to-copy GitHub/GitLab pipelines and a status.json merge gate. Governed suppressions: .greenloopignore + inline greenloop:ignore markers with mandatory reasons and expiry dates (Critical can be paused, never buried); the suppression set is itself meta-audited every round. Pre-commit-verification first for a clean baseline; invokes spec-bootstrap when no design spec exists; hands off to audit-fix at the end.
- Installs packages
3 other projects use a version of this.
Adopted by
Most recently taken up by
- spongetimblin/claude-skill-librarytheir own version, first seen Nov 8, 2026
- playbookTV/Ironcladtheir own version, first seen Nov 8, 2026
- CrowdStrike/foundry-skillstheir own version, first seen Nov 1, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- CrowdStrike/foundry-skills★ 37same name
- playbookTV/Ironclad★ 7same name
- spongetimblin/claude-skill-library★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository4 other files it gives its agents.
- audit-fixSkill
- cartographerSkill
- pre-commit-verificationSkill
- spec-bootstrapSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
78 rules. 0 are common, 50 are distinctive, and 28 are very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
50
Distinctive
This author's own choices. Worth a look.
28
Very specific
Tied to their project's names, files or tools. Probably skip.
The licence could not be identified — read it at the source. We count its rules but don't quote them. Read them on GitHub.
“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
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.