news-reaction
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a template for using ffmpeg to extract frames from video clips. This is a standard and expected operation for a video editing utility.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied footage and transcripts to generate a shot list and edit video files.
- Ingestion points: Original footage files, press screenshots, and timed transcripts provided in Step 1.
- Boundary markers: Uses a structured shot list (Step 2) to define processing boundaries.
- Capability inventory: Shell command execution via ffmpeg (Step 3) and instruction to modify local script files (video.js).
- Sanitization: Content is used for visual synchronization and frame extraction rather than direct command construction, limiting the risk of injection through transcript data.
Audit Metadata