muapi-ai-clipping
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local bash script to coordinate video processing tasks. Evidence: The SKILL.md file directs the agent to run the run-ai-clipping.sh script to interface with the clipping service.\n- [EXTERNAL_DOWNLOADS]: The skill communicates with the muapi.ai API to upload video data and retrieve processed clips. Evidence: The run-ai-clipping.sh script utilizes the muapi CLI tool to upload files and execute the clipping endpoint, representing standard interaction with the vendor's service infrastructure.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external video content, which is a potential surface for indirect prompt injection if the resulting transcriptions contain instructions for the AI agent.\n
- Ingestion points: Video content or URLs provided through the --video parameter in run-ai-clipping.sh.\n
- Boundary markers: The skill does not implement specific delimiters or safety warnings to distinguish between content and instructions in the processed results.\n
- Capability inventory: The skill uses muapi-cli for network and file operations, along with bash and jq for parsing output.\n
- Sanitization: Content returned by the API is surfaced to the agent without explicit sanitization or filtering.
Audit Metadata