tavily
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches content from external websites via the Tavily API, which may contain malicious instructions targeting the AI agent.
- Ingestion points: External web content is ingested through
scripts/search.mjs(search results) andscripts/extract.mjs(full page content). - Boundary markers: The skill uses Markdown headers to structure output but does not implement explicit boundary markers or instructions to ignore embedded commands.
- Capability inventory: The agent can execute node scripts and perform network requests.
- Sanitization: Content is returned as-is from the API without sanitization or filtering.
- [EXTERNAL_DOWNLOADS]: Communicates with the Tavily API (
api.tavily.com), which is a well-known service for AI-optimized web searching.
Audit Metadata