zoom-plugin-sdk
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references official Zoom Plugin SDK packages and documentation from
zoom.usanddevelopers.zoom.us. These are recognized as legitimate resources from a well-known service provider. - [INDIRECT_PROMPT_INJECTION]: The skill exposes a surface for indirect prompt injection by providing APIs to read untrusted data from Zoom meetings.
- Ingestion points: Meeting chat messages and participant names accessed via
Meeting.ChatandMeeting.Participantstoolkits inSKILL.mdand platform-specific references. - Boundary markers: No explicit instructions for delimiters or guardrails against embedded instructions in processed meeting data are provided.
- Capability inventory: The SDK allows for wide-ranging control over the meeting client, including participant management and media controls as described in
capabilities.md. - Sanitization: While the skill suggests checking user roles, it lacks guidance on sanitizing natural language inputs for agent-based processing.
Audit Metadata