kagi_mcp_local logo

kagi_mcp_local

by mvkvc

kagi_mcp_local is a tool intended for personal use as an alternative to manual searching and copy-pasting from Kagi Search. It allows you to run an MCP server for Kagi Search using your local browser.

View on GitHub

Last updated: N/A

What is kagi_mcp_local?

kagi_mcp_local is a local MCP (Meta-Cognitive Processor) server for Kagi Search. It allows you to query Kagi Search using your local browser as a backend, providing an alternative to manual searching and copy-pasting.

How to use kagi_mcp_local?

To use kagi_mcp_local, you need to have uv installed. The provided JSON configuration snippet shows how to configure an MCP server to use kagi_mcp_local. You'll need to configure environment variables such as BROWSER, CDP_URL, CDP_PORT, RESULTS_MAX, and PAGE_TIMEOUT to match your local setup.

Key features of kagi_mcp_local

  • Uses local browser for Kagi Search

  • Alternative to manual searching

  • Configurable environment variables

  • Personal use focused

  • MCP server implementation

Use cases of kagi_mcp_local

  • Quickly retrieve Kagi Search results programmatically for personal use

  • Automate simple search tasks without scraping

  • Integrate Kagi Search results into local applications

  • Experiment with Kagi Search MCP functionality

  • Create custom search workflows

FAQ from kagi_mcp_local

Is this tool for automated scraping?

No, this tool is intended for personal use only and not for automated querying or scraping of Kagi search results.

What are the requirements for using this tool?

You need to have uv installed.

What environment variables do I need to configure?

You need to configure BROWSER, CDP_URL, CDP_PORT, RESULTS_MAX, and PAGE_TIMEOUT.

Can I use this with any browser?

The example configuration uses Brave browser, but you can configure it to use other browsers by setting the BROWSER environment variable.

Is this tool officially supported by Kagi?

No, if you have access you should use the official Kagi MCP server at https://github.com/kagisearch/kagimcp.