The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external data that could contain adversarial instructions intended to influence the agent's behavior.
Ingestion points: Zoom meeting recordings, transcripts, and Team Chat messages (identified in SKILL.md).
Boundary markers: There are no explicit instructions for the agent to treat Zoom content as untrusted or to use specific delimiters when processing this data.
Capability inventory: The skill possesses mutation capabilities via mcp.call (mapped to zoom call-tool), including meeting_create, meeting_update, and meeting_delete (identified in SKILL.md and manifest.yaml).
Sanitization: No sanitization, filtering, or validation mechanisms for the external data are described in the skill's instructions.

zoom