youtube-video
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: Indirect prompt injection vulnerability surface detected.
- Ingestion points: Untrusted data enters the agent context through YouTube video metadata (titles, descriptions) fetched via the
--urlparameter inSKILL.md. - Boundary markers: Absent; there are no clear delimiters or instructions telling the model to ignore instructions embedded within the video content.
- Capability inventory: The skill uses
Bash(stophy *)andBash(npx @stophy/cli *)to interact with the environment. - Sanitization: No input sanitization or output validation mechanisms are specified.
Audit Metadata