The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a Python-based CLI tool (therapy-notes.py) for managing session notes. This involves local file operations (read/write/move) within a designated workspace. All operations are standard file management and do not involve shell injection risks.
[DATA_EXPOSURE]: The skill manages potentially sensitive session notes; however, these are stored locally within the user's specified workspace {WORKSPACE}/therapy-notes/. There are no network operations or external data exfiltration patterns detected.
[PROMPT_INJECTION]: The skill includes strong ethical guardrails and protocols to prevent harmful behavior. It specifically instructs the AI to maintain a professional boundary and follow crisis protocols without attempting to override system safety guidelines.

therapy-mode