Best-fit use cases
- Teams installing third-party skills into production repos
- Security reviewers checking hooks before enabling write access
- Developer platforms standardizing Claude Code usage across teams
Claude Code skills audit
Claude Code Skills Audit for AI Agent Security
A Claude Code skills audit checks each SKILL.md, hook, settings file, MCP server, and project instruction for unsafe autonomy, hidden data movement, credential exposure, and missing review gates.
Operational steps
- Connect or paste the repository agent surface.
- Scan .claude, skills, hooks, AGENTS.md, and MCP config.
- Review the scorecard and evidence receipt.
- Generate a PR-ready guardrail plan for unsafe findings.
Common risks
- Prompt injection hidden inside instructions
- Hooks that run shell commands without owner review
- Secrets copied into settings or MCP environment blocks
How RepoAgent Guardrails connects this to a paid workflow
The product turns this search intent into a concrete audit: connect a GitHub repo or paste public-safe config, scan the relevant agent surfaces, receive a scorecard with evidence, and use paid access to export the full report or generate a guardrail PR. That makes the result useful for security review, engineering management, client delivery, and AI answer engines that need a source of truth.