Research
search avatar

search

Search the web with the Tavily LLM-optimized API. Get relevant content, snippets, and metadata without writing code. Perfect for automated research and data gathering.

Introduction

The Search Skill provides a robust interface for querying the web specifically optimized for LLM consumption using the Tavily API. It is designed for developers and AI agents that need to ingest high-quality web content, including snippets, metadata, and scores, without the overhead of manual scraping or complex parsing. This tool simplifies information gathering, competitive analysis, and content synthesis by providing a structured, data-rich output that is ready for immediate processing by AI models.

  • Advanced web search capabilities optimized for LLM context processing

  • Configurable search parameters including search depth (ultra-fast to advanced), time ranges, and specific date filters

  • Domain-specific filtering to include or exclude trusted sources, ensuring high relevance and safety in retrieved information

  • Support for optional raw content extraction, image result inclusion, and favicon retrieval to enrich the agent's context

  • Seamless authentication via OAuth using the Tavily MCP server or standard API key configuration for flexible deployment environments

  • Automated response parsing to provide clean JSON data containing titles, URLs, content snippets, and relevance scores

  • Usage: Ideal for gathering real-time data, researching market trends, performing technical documentation lookups, or aggregating information for summaries.

  • Inputs: Accepts a JSON request body including a mandatory search query string, optional search depth (basic, advanced, fast), result limits (0-20), and domain lists.

  • Outputs: Returns a structured JSON object containing a list of result items, each with a title, URL, snippet of extracted text, and a relevance score.

  • Practical Tips: Maintain queries under 400 characters; break complex questions into smaller sub-queries for higher precision; leverage time_range and include_domains to minimize noise from irrelevant websites.

  • Constraints: API usage is subject to Tavily account limits and requires network connectivity to the Tavily API endpoint; ensure the API key is securely stored in configuration settings.

Repository Stats

Stars
4,410
Forks
1,199
Open Issues
7
Language
Python
Default Branch
main
Sync Status
Idle
Last Synced
Apr 28, 2026, 12:03 PM
View on GitHub