The Agent Skills Directory

[SAFE]: The skill defines a linguistic style profile and workflow for a persona agent.
- The instructions use strong language like 'CRITICAL' and 'STRICT' to enforce character consistency and language handling (English/Chinese), which is standard for high-fidelity persona skills and does not attempt to bypass core AI safety filters.
- Data retrieval is handled via specific MCP tools (retrieve_mental_models, retrieve_core_beliefs, retrieve_transcripts) with a hardcoded persona_id, which is a controlled and legitimate use of specialized tools.
- The skill does not request or use high-risk capabilities such as arbitrary command execution, network access to unknown domains, or sensitive file system operations.
- No obfuscation, persistence mechanisms, or hardcoded credentials were found.

nick-theriot