Brave Deep Research

🚀 Brave Deep Research MCP Server

A Model Context Protocol (MCP) server that combines Brave Search and Puppeteer web scraping capabilities to enable in - depth research. This server allows AI assistants not only to perform basic web searches but also to access and extract full web - page content and explore related linked pages.

🚀 Quick Start

Comparison with Standard Brave Search MCP Server

Standard Brave Search MCP Server:

• Search Capability: Performs basic web searches using the Brave Search API.
• Data Retrieval: Only returns search results (title, URL, and snippet) provided by the API.
• Content Depth: Unable to access the full web - page content beyond the search snippet.
• Page Exploration: No ability to access pages or follow links.
• Information Scope: Limited to brief information in the search results.
• Content Processing: No content extraction or cleaning functionality.
• Customization: Limited to basic search parameters (query, quantity, offset).
• Use Cases: Suitable for scenarios requiring a quick overview.

Brave Deep Research MCP Server (This Project):

• Search Capability: Uses the Brave Search API for initial results and then enhances them through web scraping.
• Data Retrieval: Extracts full page content from each search result.
• Content Depth: Provides the full content of the entire web page and extracts the main text.
• Page Exploration: Can traverse links to explore related content, with the default configuration allowing adjustable depth.
• Information Scope: Accesses comprehensive information on multiple related pages.
• Content Processing: Intelligently identifies and extracts the main content, filtering out navigation, ads, footers, etc.
• Customization: Configurable exploration depth, number of results, headless mode, and timeout settings.
• Use Cases: Suitable for in - depth research requiring detailed information and context.

Real - world Differences in Example Queries

For a query like "Climate Change Mitigation Technologies":

Standard Brave Search MCP:

Title: "Climate Change Mitigation Technologies"
URL: example.com/climate - mitigation
Snippet: Provides some technical overviews of climate change.

Brave Deep Research MCP Server:

Title: "Climate Change Mitigation Technologies"
URL: example.com/climate - mitigation
Content: Details various climate change mitigation technologies, including carbon capture, renewable energy, etc., and links to related research papers and real - world cases.
Sub - page 1: example.com/renewable - energy Provides in - depth information on renewable energy.
Sub - page 2: example.com/carbon - capture Details carbon capture technology.

📚 Documentation

Tool Parameters

The deep - search tool accepts the following parameters:

• query (Required): The search query.
• results (Optional): The number of search results to process (Default: 3, Maximum: 10).
• depth (Optional): The link traversal depth for each result (Default: 1, Maximum: 3).

💻 Usage Examples

Basic Usage

# Clone the repository
git clone https://github.com/suthio/brave-deep-research-mcp.git
cd brave-deep-research-mcp

# Install dependencies
npm install

# Run in development mode
npm run dev

# Build the project
npm run build

🔧 Technical Details

How It Works

1. The tool first uses the Brave Search API to perform a search and obtain initial results.
2. For each search result, it launches a Puppeteer browser to access the page.
3. It extracts the main content, metadata, and links from each page.
4. If the depth > 1, it traverses the links on the page and repeats the process.
5. All extracted content is formatted and returned to the AI assistant.

📄 License

MIT