video-editing
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides numerous examples of FFmpeg commands for video processing (cutting, concatenating, scaling, and normalizing audio). These are documented for the user or agent to execute locally using standard tools.
- [REMOTE_CODE_EXECUTION]: The skill includes code snippets for interacting with official APIs like ElevenLabs and fal.ai. These follow best practices, such as fetching API keys from environment variables (
os.environ["ELEVENLABS_API_KEY"]). - [DATA_EXPOSURE]: The skill mentions external services (ElevenLabs, fal.ai, Remotion), but these are well-known services within the video editing domain. It correctly instructs the use of environment variables for secrets rather than hardcoding them.
- [INDIRECT_PROMPT_INJECTION]: The skill describes processing transcripts and edit decision lists (EDLs). While this represents a data ingestion surface, the instructions are focused on structural organization and do not contain patterns that would lead to agent override.
Audit Metadata