Transcription Automation

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection, as it processes untrusted audio and video content which could contain spoken instructions designed to manipulate the AI's summarization or action-item extraction logic.
  • Ingestion points: Processes external data from audio files, video files, live streams, and meeting recordings.
  • Boundary markers: Uses structured YAML templates for output but does not explicitly implement delimiters or 'ignore' instructions to isolate processed transcripts from system prompts.
  • Capability inventory: Performs AI-driven summarization, chapter generation, and distribution to platforms like Notion and Slack.
  • Sanitization: No specific text sanitization or filtering is described for the generated transcripts before they are processed by LLMs.
  • [EXTERNAL_DOWNLOADS]: The skill facilitates data movement between several well-known third-party services and APIs as part of its core functionality.
  • Downloads content from Zoom Cloud recordings and YouTube videos.
  • Integrates with trusted AI services including OpenAI Whisper, AssemblyAI, and Deepgram for transcription.
  • Exports processed data to established productivity tools such as Slack and Notion.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 08:21 AM