The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted data from Zoom API responses (e.g., meeting topics, descriptions, and participant names).
Ingestion points: External data enters the agent context via retrieval tools such as ZOOM_LIST_MEETINGS, ZOOM_GET_A_MEETING, and ZOOM_LIST_ALL_RECORDINGS.
Boundary markers: Absent. The instructions do not define delimiters for external data or include directives for the agent to ignore instructions potentially embedded in Zoom records.
Capability inventory: The skill possesses capabilities to create, update, and delete meetings and recordings, which could be misused if the agent follows instructions found in meeting metadata.
Sanitization: Absent. There is no evidence of data sanitization or validation of content fetched from the Zoom API.
[EXTERNAL_DOWNLOADS]: The skill relies on a third-party MCP server located at https://rube.app/mcp.
This endpoint serves as the tool provider for Zoom automation, requiring the agent to send context and operation details to this external domain.
The setup instructions require users to add this URL to their MCP client configuration, introducing a dependency on a non-standard external service.

zoom-automation