Search1API MCP Server

中文文档

A Model Context Protocol (MCP) server that provides search and crawl functionality using Search1API.

Prerequisites

• Node.js >= 18.0.0
• A valid Search1API API key (See Setup Guide below on how to obtain and configure)

Installation (Standalone / General)

1. Clone the repository:

   bashCopy

   git clone https://github.com/fatwang2/search1api-mcp.git
   cd search1api-mcp
   

2. Configure API Key: Before building, you need to provide your Search1API key. See the Setup Guide section below for different methods (e.g., using a .env file or environment variables).

3. Install dependencies and build:

   bashCopy

   npm install
   npm run build
   

   Note: If using the project's .env file method for the API key, ensure it exists before this step.

Usage (Standalone / General)

Ensure your API key is configured (see Setup Guide).

Start the server:

npm start

The server will then be ready to accept connections from MCP clients.

Setup Guide

1. Get Search1API Key

1. Register at Search1API
2. Get your API key from your dashboard.

2. Configure API Key

You need to make your API key available to the server. Choose one of the following methods:

Method A: Project .env File (Recommended for Standalone or LibreChat)

This method is required if integrating with the current version of LibreChat (see specific section below).

1. In the search1api-mcp project root directory, create a file named .env:
   bashCopy

   # In the search1api-mcp directory
   echo "SEARCH1API_KEY=your_api_key_here" > .env
   

2. Replace your_api_key_here with your actual key.
3. Make sure this file exists before running npm install && npm run build.

Method B: Environment Variable (Standalone Only)

Set the SEARCH1API_KEY environment variable before starting the server.

export SEARCH1API_KEY="your_api_key_here"
npm start

Method C: MCP Client Configuration (Advanced)

Some MCP clients allow specifying environment variables directly in their configuration. This is useful for clients like Cursor, VS Code extensions, etc.

{
  "mcpServers": {
    "search1api": {
      "command": "npx",
      "args": [
        "-y",
        "search1api-mcp"
      ],
      "env": {
        "SEARCH1API_KEY": "YOUR_SEARCH1API_KEY"
      }
    }
  }
}

Note for LibreChat Users: Due to current limitations in LibreChat, Method A (Project .env File) is the required method. See the dedicated integration section below for full instructions.

Integration with LibreChat (Docker)

This section details the required steps for integrating with LibreChat via Docker.

Overview:

1. Clone this server's repository into a location accessible by your LibreChat docker-compose.yml.
2. Configure the required API key using the Project .env File method within this server's directory.
3. Build this server.
4. Tell LibreChat how to run this server by editing librechat.yaml.
5. Make sure the built server code is available inside the LibreChat container via a Docker volume bind.
6. Restart LibreChat.

Step-by-Step:

1. Clone the Repository: Navigate to the directory on your host machine where you manage external services for LibreChat (this is often alongside your docker-compose.yml). A common location is a dedicated mcp-server directory.

   bashCopy

   # Example: Navigate to where docker-compose.yml lives, then into mcp-server
   cd /path/to/your/librechat/setup/mcp-server
   git clone https://github.com/fatwang2/search1api-mcp.git
   

2. Navigate into the Server Directory:

   bashCopy

   cd search1api-mcp
   

3. Configure API Key (Project .env File Method - Required for LibreChat):

   bashCopy

   # Create the .env file
   echo "SEARCH1API_KEY=your_api_key_here" > .env
   # IMPORTANT: Replace 'your_api_key_here' with your actual Search1API key
   

4. Install Dependencies and Build: This step compiles the server code into the build directory.

   bashCopy

   npm install
   npm run build
   

5. Configure librechat.yaml: Edit your main librechat.yaml file to tell LibreChat how to execute this MCP server. Add an entry under mcp_servers:

   yamlCopy

   # In your main librechat.yaml
   mcp_servers:
     # You can add other MCP servers here too
     search1api:
       # Optional: Display name for the server in LibreChat UI
       # name: Search1API Tools
   
       # Command tells LibreChat to use 'node'
       command: node
   
       # Args specify the script for 'node' to run *inside the container*
       args:
         - /app/mcp-server/search1api-mcp/build/index.js
   

   • The args path (/app/...) is the location inside the LibreChat API container where the built server will be accessed (thanks to the volume bind in the next step).

6. Configure Docker Volume Bind: Edit your docker-compose.yml (or more likely, your docker-compose.override.yml) to map the search1api-mcp directory from your host machine into the LibreChat API container. Find the volumes: section for the api: service:

   yamlCopy

   # In your docker-compose.yml or docker-compose.override.yml
   services:
     api:
       # ... other service config ...
       volumes:
         # ... other volumes likely exist here ...
   
         # Add this volume bind:
         - ./mcp-server/search1api-mcp:/app/mcp-server/search1api-mcp
   

   • Host Path (./mcp-server/search1api-mcp): This is the path on your host machine relative to where your docker-compose.yml file is located. Adjust it if you cloned the repo elsewhere.
   • Container Path (:/app/mcp-server/search1api-mcp): This is the path inside the container. It must match the directory structure used in the librechat.yaml args path.

7. Restart LibreChat: Apply the changes by rebuilding (if you modified docker-compose.yml) and restarting your LibreChat stack.

   bashCopy

   docker compose down && docker compose up -d --build
   # Or: docker compose restart api (if only librechat.yaml changed)
   

Now, the Search1API server should be available as a tool provider within LibreChat.

Features

• Web search functionality
• News search functionality
• Web page content extraction
• Website sitemap extraction
• Deep thinking and complex problem solving with DeepSeek R1
• Seamless integration with Claude Desktop, Cursor, Windsurf, Cline and other MCP clients

Tools

1. Search Tool

• Name: search
• Description: Search the web using Search1API
• Parameters:
  • query (required): Search query in natural language. Be specific and concise for better results
  • max_results (optional, default: 10): Number of results to return
  • search_service (optional, default: "google"): Search service to use (google, bing, duckduckgo, yahoo, x, reddit, github, youtube, arxiv, wechat, bilibili, imdb, wikipedia)
  • crawl_results (optional, default: 0): Number of results to crawl for full webpage content
  • include_sites (optional): List of sites to include in search
  • exclude_sites (optional): List of sites to exclude from search
  • time_range (optional): Time range for search results ("day", "month", "year")

2. News Tool

• Name: news
• Description: Search for news articles using Search1API
• Parameters:
  • query (required): Search query in natural language. Be specific and concise for better results
  • max_results (optional, default: 10): Number of results to return
  • search_service (optional, default: "bing"): Search service to use (google, bing, duckduckgo, yahoo, hackernews)
  • crawl_results (optional, default: 0): Number of results to crawl for full webpage content
  • include_sites (optional): List of sites to include in search
  • exclude_sites (optional): List of sites to exclude from search
  • time_range (optional): Time range for search results ("day", "month", "year")

3. Crawl Tool

• Name: crawl
• Description: Extract content from a URL using Search1API
• Parameters:
  • url (required): URL to crawl

4. Sitemap Tool

• Name: sitemap
• Description: Get all related links from a URL
• Parameters:
  • url (required): URL to get sitemap

5. Reasoning Tool

• Name: reasoning
• Description: A tool for deep thinking and complex problem solving with fast deepseek r1 model and web search ability(You can change to any other model in search1api website but the speed is not guaranteed)
• Parameters:
  • content (required): The question or problem that needs deep thinking

6. Trending Tool

• Name: trending
• Description: Get trending topics from popular platforms
• Parameters:
  • search_service (required): Specify the platform to get trending topics from (github, hackernews)
  • max_results (optional, default: 10): Maximum number of trending items to return

Version History

• v0.2.0: Added fallback .env support for LibreChat integration and updated dependencies.
• v0.1.8: Added X(Twitter) and Reddit search services
• v0.1.7: Added Trending tool for GitHub and Hacker News
• v0.1.6: Added Wikipedia search service
• v0.1.5: Added new search parameters (include_sites, exclude_sites, time_range) and new search services (arxiv, wechat, bilibili, imdb)
• v0.1.4: Added reasoning tool with deepseek r1 and updated the Cursor and Windsurf configuration guide
• v0.1.3: Added news search functionality
• v0.1.2: Added sitemap functionality
• v0.1.1: Added web crawling functionality
• v0.1.0: Initial release with search functionality

License

This project is licensed under the MIT License - see the LICENSE file for details.