The Agent Skills Directory

[SAFE]: The skill is designed as a structured methodology for security analysis and does not contain any malicious code, obfuscation, or unauthorized data access mechanisms. It explicitly prohibits the agent from performing side effects such as probing production environments, modifying pull request states, or exposing sensitive data. The core checks focus on industry-standard security practices like tenant isolation, credential redaction, and proper authentication for webhooks and administrative access. No external downloads or remote code execution patterns were identified, and the use of local reference files for trust-boundary patterns is a secure and appropriate design.

review-security-privacy