zoom-rtms
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references several external libraries and services including @zoom/rtms, express, @deepgram/sdk, assemblyai, whisper-node, and openai. These are used for meeting connectivity, transcription, and AI analysis.- [DATA_EXFILTRATION]: While the skill processes sensitive meeting data (audio, video, chat), it follows security best practices by utilizing environment variables for authentication credentials (e.g., ZM_RTMS_SECRET, OPENAI_API_KEY). No evidence of unauthorized data transmission was found.- [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection because it ingests untrusted data from Zoom meetings (transcripts, chat) that may be processed by downstream AI services.
- Ingestion points: Data is ingested via client.onTranscriptData and client.onChatData as shown in examples/ai-integration.md and examples/sdk-quickstart.md.
- Boundary markers: The provided examples do not demonstrate the use of delimiters or 'ignore' instructions for the ingested meeting content.
- Capability inventory: The skill facilitates network operations (WebSocket, HTTPS) and local file writing.
- Sanitization: No explicit sanitization of transcript or chat data is implemented in the provided AI integration snippets.
Audit Metadata