The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions to use curl commands via Bash to interact with administrative endpoints (e.g., http://localhost:3000/api/admin/refresh/x-posts). These commands are designed to trigger internal data synchronization processes.
[SAFE]: The skill mentions checking environment files (.env.local) for XAI_API_KEY and YOUTUBE_API_KEY. This is documented as a troubleshooting step for administrators to ensure correct configuration and is not an instruction for credential exfiltration.
[SAFE]: The skill outlines a data flow involving external content ingestion from X and YouTube. While this presents an indirect injection surface, the skill is focused on the administrative triggering of the sync and mentions existing sanitization logic in the project source code.
Ingestion points: External data enters the system from XAI and YouTube APIs.
Boundary markers: No specific delimiters or boundary instructions are included in the skill's operational steps.
Capability inventory: The skill's environment allows Bash and Read tools.
Sanitization: The instructions specifically reference src/lib/xai.ts for handling and sanitizing potentially malformed JSON from external sources.

refresh-content