websearch

WebSearch MCP Server provides a multi-tool web search framework with built-in capabilities to perform web searches across multiple providers, convert web pages to Markdown, retrieve content from WeChat official account articles, and perform GitHub searches. Once running, you can leverage these features via the server's endpoints (documented in the project documentation). The server supports providers like Bing, Baidu, and Google (when browser_use is enabled), and includes a Markdown converter that turns URL content into Markdown format. It also includes a WeChat account article retriever and a convenient GitHub search interface for repositories, users, issues, and more. To operate, start the server with the recommended Python/uv workflow: you can run the server with uv or directly with Python after activating the virtual environment. After startup, configure your environment (.env) to tailor the search behavior (e.g., language, region, and host/port).

Prerequisites: - Python 3.12 or higher - uv for dependency management (or pip) Step-by-step: 1. Clone the repository:

git clone https://github.com/Howe829/websearch-mcp-server.git
cd websearch-mcp-server
``` 2. Set up dependencies (recommended via uv): 
```bash
uv venv && uv sync

Or install via pip from requirements.txt:

pip install -r requirements.txt
``` 3. Create and configure environment: 
```bash
cp .env.example .env

Edit .env to set BING_SEARCH_BASE_URL, LANGUAGE, CC, IMPERSONATE, HOST, PORT, SERVER_MODE as needed. 4. Run the server:

uv run python server.py

or simply:

source .venv/bin/activate
python server.py

Tips and notes: - The server relies on a .env file for configuration. Copy .env.example to .env and customize values like BING_SEARCH_BASE_URL, LANGUAGE, CC, IMPERSONATE, HOST, PORT, and SERVER_MODE. - If you prefer pip, you can install dependencies with pip install -r requirements.txt but uv is the recommended workflow in this project. - When testing, you can run tests with pytest via uv: uv pip install pytest followed by uv run pytest. - Code style can be checked with Ruff: uv pip install ruff and ruff format .. - For debugging MCP integration, you can use the MCP inspector: uv run fastmcp dev server.py. - Ensure your environment provides network access to external search providers and that any API keys or access tokens required by providers are configured in the .env file if needed.