The Agent Skills Directory

[SAFE]: The skill contains purely educational content and templates designed to guide users through development workflows. No malicious code, automated exploits, or dangerous commands were detected.- [CREDENTIALS_UNSAFE]: The skill includes instructions on managing environment variables and secrets. It uses dummy placeholders (e.g., 'your_key_here', 'abc123...') for demonstration purposes and explicitly advises users on how to prevent credential leakage using .gitignore and key revocation.- [EXTERNAL_DOWNLOADS]: Mentioned packages (e.g., 'claude-mpm') and platforms (e.g., Vercel, Netlify) are standard industry tools used for instructional context. There are no automated downloads or remote script executions performed by the skill itself.- [DATA_EXFILTRATION]: No network operations or data harvesting patterns were found. The skill focuses on structured teaching moments within the agent's conversation flow.- [PROMPT_INJECTION]: The skill defines 'Circuit Breakers' which are internal constraints used to simulate a teaching persona. These do not attempt to bypass or override the underlying system safety guidelines.

pm-teaching-mode