video-editing
Pass
Audited by Gen Agent Trust Hub on Mar 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides various FFmpeg command templates for tasks like video trimming, concatenation, and audio normalization.
- [DATA_EXFILTRATION]: Includes Python examples for using the ElevenLabs API, correctly suggesting environment variables for credential management.
- [PROMPT_INJECTION]: Identifies an indirect prompt injection surface where the agent processes transcripts from raw footage to plan video edits. 1. Ingestion points: Raw footage transcripts (Layer 2 in SKILL.md). 2. Boundary markers: No delimiters recommended for input data. 3. Capability inventory: Command execution (FFmpeg), file writing, and network API calls (ElevenLabs, fal.ai). 4. Sanitization: No transcript sanitization or validation is mentioned.
Audit Metadata