media-transcription-report

Pass

Audited by Gen Agent Trust Hub on Oct 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill relies on executing various shell commands to perform media processing and document compilation.
  • Evidence: Usage of ffmpeg for audio extraction, ffprobe for inspection, and pdflatex/xelatex for PDF generation.
  • Evidence: Use of conda run to execute whisperx and whisper transcription engines.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data that could contain malicious instructions designed to influence the report generation phase.
  • Ingestion points: The skill ingests transcripts generated from user-provided audio/video files and reads chat/source notes (SKILL.md).
  • Boundary markers: Includes instructions to maintain a "neutral third-person" style and explicitly forbids mentioning the chat context or AI-generation process in output reports.
  • Capability inventory: The skill has the ability to write to the file system, execute media processing binaries, and compile LaTeX documents (SKILL.md).
  • Sanitization: Relies on structural formatting requirements and tone constraints to isolate the final report from potentially malicious content in the source material.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 19, 2026, 06:57 PM