Audit Integrity Checks
Makes an agent verify security findings with code evidence and score reports before delivery
- Repository
- github/awesome-copilot
- Stars
- 40,703
- Licence
- MIT
- Files
- 8 files
What it does
This is a set of quality rules for an AI agent doing security reviews, such as static analysis, dependency checks, threat modelling or code reviews. Once installed, the agent is asked to verify each finding against actual code locations and data-flow traces, avoid dismissing valid findings, ask at most two clarifying questions, retry failed tool calls once before documenting the gap, run a second self-critique pass, and score its draft report against a 1-10 rubric with a minimum of 8 before delivering it. It also asks the agent to record lessons learned after a scan.
Great for
- Developers or security engineers who use an AI agent to run security audits and want more rigorous, evidence-backed reports
- Teams that want a consistent quality checklist applied to AI-generated code review output
Skip it if
- Simple one-off code questions where a full audit workflow is unnecessary overhead
- Anyone expecting the skill to actually find vulnerabilities by itself: it governs output quality and does not bundle scanners or security tools
- Situations needing a guaranteed-correct audit: the scoring is self-assessed by the same agent, so it can still be wrong
Try asking
- “Review this Node.js service for injection vulnerabilities and make sure every finding has a code location and taint trace”
- “Threat model this payment API and check the draft report against the quality gate before giving it to me”
- “Run a dependency audit on this repo and document any tool failures instead of assuming the packages are safe”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 8 original files, the MIT licence and an attribution note.
audit-integrity.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip audit-integrity.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip audit-integrity.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Provide the codebase or scan output in the conversation so the agent can actually verify code locations and traces
- Pair it with a real security tool or scanner, since this skill governs reasoning quality rather than detecting issues
- Treat the self-assigned quality score as a prompt for review, not as independent validation
What's inside
- references/anti-rationalization-guard.md
- references/clarification-protocol.md
- references/non-negotiable-behaviors.md
- references/retry-protocol.md
- references/self-critique-loop.md
- references/self-learning-system.md
- references/self-reflection-quality-gate.md
- SKILL.md
Redistributed unmodified from github/awesome-copilot under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.