Verified

Omnisearch MCP Server

Tavily, Brave, Kagi, Exa, GitHub, Linkup, and Firecrawl unified in four search tools.

Data last scanned today · Reviewed today

454 tools★ 348TypeScriptMIT

Overview

Omnisearch MCP Server streamlines AI research by consolidating access to seven distinct search and data extraction services into a single interface. By providing four specialized tools—web_search, ai_search, github_search, and web_extract—it enables developers to pull data from diverse providers like Kagi, Tavily, and Firecrawl directly through their AI assistant. Configuration is flexible, allowing users to provide specific API keys for the services they intend to use, while automatically skipping those without configured credentials. This modular approach permits integration via MCPick or manual configuration, offering options to run the server from source or use the published package.

Our verdict

With a security score of 40/100, this server is classified as high-risk. This rating reflects its community-maintained status and the fact that its authentication methods and read-only mode compliance have not yet been editorially reviewed. Users should exercise caution and manually audit the implementation before integrating it into environments requiring strict data security protocols.

Consolidates seven search and extraction services into one interface
Contains only five direct dependencies to minimize supply chain surface
Actively maintained with the latest commit occurring today
Includes documented usage examples for all four provided tools

Community-maintained project lacking official vendor oversight
Authentication methods have not been editorially reviewed
Read-only mode support remains unverified for safety
No official homepage or website link provided

Setup

Claude Desktop

{
  "mcpServers": {
    "mcp-omnisearch": {
      "args": [
        "-y",
        "mcp-omnisearch"
      ],
      "command": "npx"
    }
  }
}

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.

No read-only mode — running this grants full read/write access.
No built-in authentication — anyone with access to this config can invoke every tool it exposes.

Tools

Tool	Description	Risk
web_search	Search the web with Tavily, Brave, Kagi, Exa, or Kagi Enrichment.	low
ai_search	Get sourced AI answers with Kagi FastGPT, Exa Answer, Linkup, or Tavily Research.	low
github_search	Search GitHub code, repositories, or users.	low
web_extract	Extract, crawl, scrape, summarize, or find similar content with Tavily, Kagi, Firecrawl, or Exa.	low

Compatibility

Client	Local	Docker	Remote	Read-only
ChatGPT
Claude Desktop
Cursor
VS Code
Windsurf

Frequently asked questions

›Do I need an API key for every supported search provider?

No, you only need to provide keys for the providers you intend to use. The server will skip any providers that do not have a configured API key.

›How do I configure the server if I am not using MCPick?

You can manually configure the server by adding it to your MCP configuration file and providing your API keys as environment variables.

›Does the server support self-hosted Firecrawl?

Yes, you can specify a custom endpoint by setting the FIRECRAWL_BASE_URL environment variable.

›How does the server handle large search results?

The server defaults to file-based output for large results, but you can change this behavior by setting the OMNISEARCH_LARGE_RESULT_MODE environment variable to inline.

›How do I retrieve a Tavily Research report?

The ai_search tool returns a task ID for Tavily research; you must pass this ID back to the tool as a research_id to retrieve the final report.