ffmpeg-docs

Pass

Audited by Gen Agent Trust Hub on Jun 27, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches documentation from the official FFmpeg project website.
  • Evidence: The script scripts/ffdocs.py contains a hardcoded BASE URL pointing to https://ffmpeg.org and uses urllib.request to retrieve content.
  • [COMMAND_EXECUTION]: Instructs the agent to execute a local Python script for documentation lookup and management.
  • Evidence: SKILL.md contains multiple examples demonstrating how to use uv run to execute scripts/ffdocs.py for search, fetch, and indexing tasks.
  • [PROMPT_INJECTION]: The skill ingests data from an external website, which creates an indirect prompt injection surface.
  • Evidence: The fetch and search commands in scripts/ffdocs.py provide external documentation content to the agent.
  • Ingestion points: scripts/ffdocs.py reads HTML content from ffmpeg.org manually.
  • Boundary markers: The script provides structure by identifying headings and anchor IDs in the output.
  • Capability inventory: The skill facilitates network read operations, local file writes for caching, and script execution.
  • Sanitization: The script uses a custom TextExtractor class to parse HTML, explicitly skipping <script>, <style>, <nav>, and <footer> tags to provide clean text to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 27, 2026, 09:54 AM
Security Audit — agent-trust-hub — ffmpeg-docs