The Agent Skills Directory

[SAFE]: The skill serves as a legitimate technical resource for security professionals. It documents the use of industry-standard security tools such as Semgrep, Bandit, Gitleaks, Trivy, and nmap.
[COMMAND_EXECUTION]: While the skill references numerous shell commands for scanning and testing, these are provided as a library of reference materials for the agent to use during authorized security audits. The instructions include strict constraints prohibiting testing on production systems or testing without explicit written authorization.
[PROMPT_INJECTION]: No attempts to override agent behavior or bypass system safety filters were identified. The instructions use standard pedagogical language to guide the agent through a security review process.
[DATA_EXFILTRATION]: No patterns associated with credential theft or unauthorized data exfiltration were found. The 'Secret Scanning' reference file provides patterns used to detect hardcoded secrets within a codebase rather than attempting to harvest them.
[EXTERNAL_DOWNLOADS]: The skill mentions installing standard security packages from well-known registries (e.g., pip install bandit, npm install eslint-plugin-security). These are industry-standard practices for setting up a security analysis environment.

security-reviewer