remotion-vlog
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires the agent to execute system commands for media file verification (SHA-256), metadata inspection (ffprobe), and hardware volume management (ejecting mounted volumes). These are standard operations for media handling.\n- [EXTERNAL_DOWNLOADS]: The agent is instructed to download royalty-free music and retrieve location data from official sources like OpenStreetMap. These downloads are functional requirements and do not involve untrusted code execution.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external websites and media metadata, which constitutes an attack surface. The risk is mitigated by specific editorial instructions and a mandatory preview approval step.\n
- Ingestion points: Web-based location research, media metadata, and user-provided reference images (SKILL.md).\n
- Boundary markers: No explicit delimiters for external data, but the workflow includes a human-in-the-loop preview gate.\n
- Capability inventory: File system read/write, network access, and shell command execution.\n
- Sanitization: Relies on the agent's adherence to editorial guidelines and manual verification of rendered stills.
Audit Metadata