video-editing
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill facilitates a workflow susceptible to indirect prompt injection.
- Ingestion points: Untrusted external data in the form of video transcripts and raw footage metadata is processed in Layer 2.
- Boundary markers: No explicit delimiters or safety instructions are provided to the agent to distinguish between its own instructions and the content of the transcripts.
- Capability inventory: The skill allows for the execution of shell commands (FFmpeg, Remotion), file system writes, and network operations via API calls.
- Sanitization: There is no requirement or guidance for the agent to sanitize or validate transcript data before using it to generate shell commands.
- [COMMAND_EXECUTION]: The skill relies on the generation and execution of system-level commands to perform its tasks.
- Evidence: The skill provides templates for FFmpeg bash scripts in Layer 3 and uses 'npx remotion' for video rendering in Layer 4.
Audit Metadata