tavily-due-diligence

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (low risk: 0.10). The workflow requires searching and extracting content from arbitrary public web sources, including news, court dockets, regulatory filings, and company statements via Tavily search.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the agent to execute an external installation command (npx skills add tavily-ai/skills --skill <name>) to fetch and load runtime skill material from an unverified GitHub repository (tavily-ai/skills), introducing a high-tier unverifiable operational dependency.