The Agent Skills Directory

[SAFE]: The skill is designed for internal reasoning ('internal mode') and user interaction ('user-facing mode') to facilitate decision making. It explicitly states an invariant that it writes no production code, no specs, and does not commit changes or mutate project state, focusing solely on knowledge recording. Findings show it delegates file writing (wheypoints and ledgers) to specific tools like /wheypoint and uses markdown manipulation for documentation purposes. No malicious command execution, unauthorized data access, or obfuscation was detected. Its primary function is to model problems and decide on next steps within a pipeline (mold, cook, press, etc.).

culture