The Agent Skills Directory

[COMMAND_EXECUTION]: The skill interacts with Zoom services by executing the local zoom CLI binary for authentication, status checks, and tool calls.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external content such as meeting transcripts, recordings, and chat messages, creating a surface for indirect prompt injection. • Ingestion points: Reviewing meeting transcripts, recordings, and Team Chat via the zoom CLI. • Boundary markers: None are explicitly defined for separating external data from agent instructions. • Capability inventory: The skill can perform write operations including meeting_create, meeting_update, and meeting_delete. • Sanitization: No explicit sanitization is described, but the skill mitigates risk by requiring explicit user approval for every tool call and write operation.

zoom