tavily-best-practices
Pass
Audited by Gen Agent Trust Hub on Feb 26, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of Markdown documentation and code examples for developers. No executable scripts, prompts, or active code components are included within the skill files.
- [EXTERNAL_DOWNLOADS]: The documentation references official packages and tools such as tavily-python, @tavily/core, and langchain-tavily from well-known registries. These are standard dependencies for the Tavily service and its integrations.
- [COMMAND_EXECUTION]: The skill includes setup examples for installing dependencies and configuring environment variables (such as TAVILY_API_KEY), which is standard and expected behavior for technical documentation.
- [DATA_EXFILTRATION]: All network-related documentation points to official Tavily API endpoints (api.tavily.com and mcp.tavily.com). No suspicious or unauthorized destinations are identified in the provided examples.
Audit Metadata