transcribe
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on the execution of shell commands using yt-dlp, mv, and the MacWhisper CLI (mw) to download, manage, and transcribe media files. While these are standard tools, the use of shell commands with external inputs like URLs presents a standard attack surface.- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves processing external media content from user-provided URLs to generate text transcripts, which creates a potential surface for indirect prompt injection attacks.\n
- Ingestion points: Remote video and audio URLs processed in SKILL.md.\n
- Boundary markers: The instructions lack explicit delimiters or warnings to the agent to ignore any instructions that might be contained within the generated transcript text.\n
- Capability inventory: The skill uses shell execution for media downloading (yt-dlp), file movement (mv), and local transcription (mw) as defined in SKILL.md.\n
- Sanitization: There is no mechanism described to sanitize the textual output of the transcription process before it is consumed by the agent.
Audit Metadata