SuperAPI AI Directory

g-search-mcp

Google search with parallel queries for multiple keywords.

Open website

About this tool

Purpose

g-search-mcp is an MCP (Model Context Protocol) server for Google Search, mainly addressing scenarios that require searching with multiple keywords at the same time. It allows submitting multiple query terms at once and executing searches in parallel, thereby improving search efficiency. This project is modified from google-search.

Main Capabilities

  • Parallel Search: Supports searching on Google with multiple keywords at the same time, improving search efficiency.
  • Browser Optimization: Opens multiple tabs in a single browser instance to achieve efficient parallel search.
  • Automatic Verification Handling: Intelligently detects CAPTCHA, and enables visible browser mode when needed for the user to complete verification.
  • User Behavior Simulation: Simulates real user browsing patterns to reduce the likelihood of being detected by search engines.
  • Structured Data: Returns structured search results in JSON format, making them easy to process and analyze.
  • Configurable Parameters: Supports various parameter configurations such as the maximum number of search results, timeout settings, and regional language settings.

Usage

Run directly via npx

During initial setup, you need to run a command in the terminal to install the required browser.

You can use the --debug option to run in debug mode (showing the browser window).

Configure in Claude Desktop

Configuration file locations:

  • MacOS:~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows:%APPDATA%/Claude/claude_desktop_config.json

search tool

Executes multi-keyword Google searches and returns results, using the Playwright browser for searching. Supports the following parameters:

  • queries: Array of search queries to execute (required parameter)
  • limit: Maximum number of results returned per query, default 10
  • timeout: Page load timeout in milliseconds, default 60000 (60 seconds)
  • noSaveState: Whether to avoid saving browser state, default false
  • locale: Language and region setting for search results, default en-US
  • debug: Whether to enable debug mode (showing the browser window), which overrides the --debug flag in the command line

Parameter Adjustment Examples

  • Increase the number of search results: For example, if you request "return the top 20 search results for each keyword", the limit: 20 parameter will be set.
  • Increase the timeout: For slow loading situations, for example, "set the page load timeout to 120 seconds", the timeout parameter will be adjusted to 120000 milliseconds.
  • Change the search region: Specify a different language region, for example, "please search using the Chinese region (zh-CN)", and the locale: "zh-CN" parameter will be set.

Debugging and Troubleshooting

  • Dynamically enable debugging: Show the browser window during a specific search operation, for example, "please enable debug mode for this search operation"; even if the --debug flag was not added when the server started, debug: truewill still be set.

Environment Requirements and Installation

  • Requires Node.js 18 or higher.

Install from Source

  1. Clone the repository
  2. Install dependencies
  3. Install the Playwright browser
  4. Build the server

Automatic Rebuild (Development Mode)

Supports automatic rebuild in development mode.

Debugging with MCP Inspector

MCP Inspector can be used for debugging.

Related Projects

  • fetcher-mcp: An MCP server that uses the Playwright headless browser to fetch web page content, with features such as intelligent content extraction, parallel processing, and resource optimization, suitable for web content scraping.

License

This project is licensed under the MIT License.