video-editing
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides numerous shell commands for using FFmpeg to process video files, including cutting, concatenating, and reframing footage. These are standard operations for the tool's intended purpose.
- [EXTERNAL_DOWNLOADS]: The skill interacts with well-known services, specifically the ElevenLabs API for speech synthesis and Remotion (via npx) for programmable video composition. These references represent normal usage of established third-party platforms.
- [INDIRECT_PROMPT_INJECTION]: The workflow involves processing user-provided transcripts and recordings. Ingestion points: Claude/Codex analysis of raw recordings (SKILL.md). Capability inventory: Subprocess calls to FFmpeg, file writing for audio/video output, and network requests to the ElevenLabs API. No specific boundary markers or sanitization steps are defined for the input transcripts.
- [DYNAMIC_EXECUTION]: The skill instructs the agent to dynamically generate FFmpeg commands and Remotion React components based on transcript analysis. This is a core feature of the workflow and uses template-based generation.
Audit Metadata