Security Audit skill
Review trust boundaries with independent verification and structured findings.
Official · Cloudflare| Author | Cloudflare |
|---|---|
| Licence | MIT |
| Source | cloudflare/security-audit-skill · SKILL.md on GitHub |
| Works in | Claude Code · Codex, Gemini CLI, Cursor, Copilot |
| Read from the source on |
The security-audit skill coordinates source-first vulnerability reviews across six phases.
Guidance mode answers focused questions; an explicit full audit produces validated records and reports.
From the skill's SKILL.md: ## Core principles
What it does
- Maps boundaries, hunts candidates and assigns fresh verifiers.
What you get
- findings.json, coverage-ledger.json and Markdown reports.
When to use it
- Review a repository you are authorized to audit.
When not to use it
- A guarantee that software has no vulnerabilities.
What it needs
- Node.js, tool-using subagents and an OS-enforced sandbox for target execution.
Files in the skill
| File or folder | What it is for |
|---|---|
SKILL.md | Entry-point instructions and workflow routing. |
report-schema.json | Schema separating confirmed, unresolved and rejected records. |
Licence as found in the repository on 2026-11-16: MIT.
Example prompts
Type one of these after installing; naming the skill makes sure it is used.
- Security audit this repository.
- Review this authentication boundary.
- Explain unresolved findings in this report.
How to install security-audit
Prepare the workflow
Keep the complete skill folder. Full execution needs isolated networking, an allowlisted environment, scratch-only writes and resource limits. Without them, retain unresolved validation.
node ~/.claude/skills/security-audit/validate-findings.cjs findings.jsonA skill is a folder with a SKILL.md file. You install it once; the agent reads its short description at the start of each session and loads the full instructions only when a request matches. Pick the place you use:
Claude Code: copy the folder into your skills folder
Put the skill's folder in ~/.claude/skills/ to use it in every project on this computer, or in .claude/skills/ inside a repository to share it with everyone who works there (commit it).
git clone --depth 1 https://github.com/cloudflare/security-audit-skill.git
mkdir -p ~/.claude/skills
cp -R security-audit-skill/skills/security-audit ~/.claude/skills/Claude Code notices new or changed skills in these folders during the session. If the skills folder did not exist when the session started, run /reload-skills once. Then type /security-audit or just describe the task.
On Windows, run the commands in Git Bash or WSL, or copy the folder into the same place with File Explorer.
Codex, Gemini CLI, Cursor and GitHub Copilot
These agents read the same SKILL.md folders. The one folder all four accept inside a project is .agents/skills/, so this copies the skill there:
git clone --depth 1 https://github.com/cloudflare/security-audit-skill.git
mkdir -p .agents/skills
cp -R security-audit-skill/skills/security-audit .agents/skills/| Agent | Project folder | Your user folder | Notes |
|---|---|---|---|
| OpenAI Codex | .agents/skills/ | ~/.agents/skills/ | Type $ or run /skills to pick a skill; Codex also chooses one when your task matches. |
| Gemini CLI | .agents/skills/.gemini/skills/ | ~/.agents/skills/~/.gemini/skills/ | Or install from Git with gemini skills install <git URL> --path <folder> --consent. Google replaced Gemini CLI with Antigravity CLI for free-tier and Google One users on 18 June 2026; Antigravity uses .agents/skills/ in a project and ~/.gemini/config/skills/ for you. |
| Cursor | .agents/skills/.cursor/skills/ | ~/.agents/skills/~/.cursor/skills/ | Type / in Agent chat to pick a skill. Cursor also reads Claude Code's .claude/skills/ folders. |
| GitHub Copilot | .github/skills/.agents/skills/.claude/skills/ | ~/.copilot/skills/~/.agents/skills/ | Works in Copilot CLI, the cloud agent, and agent mode in VS Code, Visual Studio and JetBrains (IDE support depends on the version). Or run gh skill install OWNER/REPOSITORY SKILL (public preview, GitHub CLI 2.90.0 or later). |
Sources for these steps
Checked against each vendor's documentation on 2026-11-14:
Coverage is evidence
The ledger tracks inspected units and gaps. Repeated runs can revisit changed source and prior unresolved leads.
Separate certainty from severity
Only confirmed records receive severity. Missing deployment facts remain needs_validation rather than assumed vulnerabilities.
Questions about security-audit
Does loading it start an audit?
No. Full audits require an explicit request.
Is every candidate a vulnerability?
No. Fresh verification can reject a candidate.
Does it fix source code?
It recommends narrow fixes without modifying target source.
Is this a live website scanner?
No. It emphasizes source and bounded local evidence.