media-fixing-and-repair
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill includes a dedicated security section that explicitly instructs the AI to treat all metadata extracted from media files as untrusted external data. It warns the agent not to follow any instructions found within these fields, which is an excellent security control against indirect prompt injection.
- [COMMAND_EXECUTION]: The skill provides numerous FFmpeg and FFprobe command templates. These are standard media processing commands and do not involve privilege escalation or unauthorized file access.
- [PROMPT_INJECTION]: A static analysis flag for prompt injection was found to be a false positive; the text in question is actually a security rule directing the agent to ignore malicious instructions found in untrusted data.
Audit Metadata