The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external, untrusted content such as essays, scripts, and video descriptions, which can contain hidden instructions aimed at altering the agent's behavior. \n- Ingestion points: The skill ingests user-authored artifacts for analysis as described in the 'Read the Whole Piece' section of SKILL.md. \n- Boundary markers: The instructions do not define delimiters or provide warnings for the agent to ignore instructions embedded within the analyzed content. \n- Capability inventory: The skill is limited to providing a textual evaluation and does not utilize tools for network access, file system manipulation, or command execution. \n- Sanitization: No logic is provided to sanitize or filter the input data for malicious prompt segments.

publication-pass