revenue-maxxing
Pass
Audited by Gen Agent Trust Hub on Nov 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which can be used as an attack vector to influence the agent's behavior or inject malicious content into the final output.
- Ingestion points: The workflow instructs the agent to read YouTube 'About' pages, video descriptions, transcripts, and mine viewer comments for specific keywords and 'pains'.
- Boundary markers: The instructions lack explicit boundary markers (e.g., XML tags or delimiters) to separate untrusted external content from the agent's instructions, nor do they include warnings to ignore embedded commands within the fetched data.
- Capability inventory: The skill has the capability to generate HTML files (artifacts) and perform searches via the 'TubeLab MCP'.
- Sanitization: There is no evidence of sanitization or escaping of the content mined from transcripts or comments before it is interpolated into the HTML template (e.g., in the '{a viewer's words}' placeholder), which could lead to cross-site scripting (XSS) if the generated report is viewed in a browser.
Audit Metadata