ffmpeg-audio-processing
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides numerous examples of shell commands and bash loops intended for local execution. These commands use the FFmpeg utility for standard media manipulation tasks such as conversion, compression, and audio filtering.
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the processing of external data by instructing the agent to read metadata from media files using the
ffmpeg -icommand. This creates a potential surface for indirect prompt injection via malicious metadata tags. 1. Ingestion points: Metadata extraction from user-provided audio and video files (e.g., SKILL.md sections for viewing properties and extracting audio). 2. Boundary markers: Absent; there are no instructions to the agent to disregard content within the tool output. 3. Capability inventory: The skill utilizes the FFmpeg binary for file system reading, file writing, and command execution. 4. Sanitization: Absent; there is no validation or filtering performed on the metadata output before it is processed by the agent.
Audit Metadata