The Agent Skills Directory

[SAFE]: No security issues detected. The skill includes a dedicated 'Security Validation' framework (documented in references/security-checks.md) that explicitly instructs the agent to perform self-validation before saving files. These checks cover:
- Path Traversal: Explicitly rejects patterns like ../, system directories (/etc, /var), and sensitive user paths (.ssh, .aws).
- Secrets Detection: Provides patterns to identify and replace hardcoded credentials (API keys, tokens, passwords) with placeholders.
- Injection Protection: Identifies potential SQL, shell, and code injection patterns in verification criteria.
[SAFE]: The skill's workflow (defined in SKILL.md) follows a multi-step confirmation process that prevents unauthorized file modifications by requiring explicit user approval before proceeding with core tasks or overwriting existing documentation.

req-traceability