fastify-code-review
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted pull request data and executes repository scripts, creating a surface for malicious code execution if the PR content contains harmful script definitions. Ingestion points include PR metadata and diffs retrieved via the GitHub CLI. Boundary markers include instructions to use isolated checkouts and draft reviews locally. Capability inventory includes git, gh, and npm execution. Sanitization involves explicit instructions to inspect script bodies and lifecycle hooks before execution.
- [COMMAND_EXECUTION]: The skill relies on system commands using git, gh, and npm to perform review tasks and verification, which involve executing logic defined in the target repository.
- [DYNAMIC_EXECUTION]: The skill performs dynamic execution of repository-defined scripts and tools such as astro, biome, and action-validator, and handles scripts that may regenerate files (e.g., test:validator:integrity) within isolated environments.
Audit Metadata