The Agent Skills Directory

[SAFE]: The skill serves as a documentation and methodology resource for repository auditing. It instructs the agent to operate in a read-only capacity, focusing on identifying architectural smells, quality issues, and security vulnerabilities within a project's source code.
[SAFE]: The Python utility scripts/validate-frontend-visual-evidence.py is a self-contained JSON schema validator. It does not utilize any third-party libraries (relying on json, hashlib, base64, etc.) and does not perform any network requests or dangerous system operations. It contains proper path traversal protections (e.g., checking for /../ in artifact locators).
[SAFE]: No evidence of prompt injection, obfuscation, or malicious command execution was found. The skill follows best practices by explicitly defining boundaries and requiring evidence-based findings for code quality claims.
[SAFE]: All external references and corpus links target legitimate, well-known open-source projects (e.g., shadcn/ui, Tokio, ripgrep) and do not involve suspicious domains or unverifiable remote code sources.

repo-audit