The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from social media comments via tools like list_replies and open_post to draft replies, creating a potential surface for malicious input to influence agent behavior. \n- Ingestion points: External social media comments and replies retrieved via API tools. \n- Boundary markers: The instructions do not define specific delimiters to separate external data from agent instructions. \n- Capability inventory: The skill has the ability to post replies, hide comments, and delete content using the reply, hide_reply, and delete_reply tools. \n- Sanitization: No explicit sanitization of input text is described, though the requirement for explicit human approval before any action serves as a primary security control.

comments