MCP Documentation Search Server

A comprehensive Model Context Protocol (MCP) server that provides powerful search and browsing capabilities for MCP documentation and Python SDK content. This server enables efficient exploration of MCP concepts, implementation details, and code examples through intelligent parsing and search functionality.

🚀 Features

📚 Resource Browsing

• Section Listing - Browse all available documentation sections with metadata
• Full Section Content - Access complete content of any documentation section
• Core Concepts Overview - Quick access to fundamental MCP concepts

🔍 Advanced Search Tools

• Section Name Search - Find sections by title and header names
• Content Text Search - Full-text search within documentation content
• Implementation Requirements - Locate imports, dependencies, and setup instructions
• Contextual Results - Search results include surrounding context for better understanding

📊 Rich Metadata

• Document type classification (MCP Documentation vs Python SDK)
• Header level hierarchy (H1, H2, H3, etc.)
• Word count statistics
• Match frequency analysis

📦 Installation

Prerequisites

• Python 3.8 or higher
• MCP Python SDK

Dependencies

pip install "mcp[cli]"

Quick Setup

1. Save the server file as mcp_docs_server.py
2. Make it executable: chmod +x mcp_docs_server.py

🏃 Running the Server

Development Mode

Test and debug your server with the MCP Inspector:

mcp dev mcp_docs_server.py

Claude Desktop Integration

Install the server in Claude Desktop for seamless integration:

mcp install mcp_docs_server.py --name "MCP Documentation Search"

Direct Execution

Run the server directly:

python mcp_docs_server.py

🔧 API Reference

Resources

mcp://sections/list

Get a comprehensive list of all available documentation sections.

Returns: Hierarchical list of sections with metadata including:

• Section names and titles
• Document type (MCP Documentation/Python SDK)
• Header level hierarchy
• Word count for each section

mcp://section/{section_name}

Retrieve the complete content of a specific documentation section.

Parameters:

• section_name (string): Exact name of the section to retrieve

Returns: Full section content with metadata

mcp://concepts/core

Access an overview of core MCP concepts including servers, resources, tools, and prompts.

Returns: Curated overview of fundamental MCP concepts with examples

Tools

search_sections_by_name(keyword, case_sensitive=False, max_results=10)

Search for documentation sections by section names containing the specified keyword.

Parameters:

• keyword (string): The keyword to search for in section names
• case_sensitive (boolean): Whether to perform case-sensitive search (default: False)
• max_results (integer): Maximum number of results to return (default: 10)

Returns: List of matching sections with previews and metadata

Example:

# Find all sections related to FastMCP
search_sections_by_name("FastMCP")

# Case-sensitive search for tools
search_sections_by_name("Tools", case_sensitive=True, max_results=5)

search_content_by_text(keyword, case_sensitive=False, max_results=10)

Search for sections containing the keyword in their content text.

Parameters:

• keyword (string): The keyword to search for in content text
• case_sensitive (boolean): Whether to perform case-sensitive search (default: False)
• max_results (integer): Maximum number of results to return (default: 10)

Returns: List of matching sections with contextual snippets around matches

Example:

# Find content about tool decorators
search_content_by_text("@mcp.tool")

# Search for import statements
search_content_by_text("import", max_results=15)

get_section_details(section_name)

Get the complete content of a specific documentation section.

Parameters:

• section_name (string): The exact name of the section to retrieve

Returns: Full section content with comprehensive metadata

Example:

# Get details about core concepts
get_section_details("Python SDK: Core Concepts")

find_implementation_requirements(keyword)

Find implementation requirements, imports, and dependencies for a specific MCP feature or concept.

Parameters:

• keyword (string): The feature or concept to find requirements for

Returns: Requirements, imports, and implementation details with code examples

Example:

# Find FastMCP implementation requirements
find_implementation_requirements("FastMCP")

# Get server setup requirements
find_implementation_requirements("server")

📖 Usage Examples

Basic Search Operations

# Browse available sections
sections = get_resource("mcp://sections/list")

# Search for authentication-related content
auth_sections = search_sections_by_name("auth")

# Find all mentions of decorators in content
decorator_content = search_content_by_text("@mcp")

# Get complete implementation details
server_details = get_section_details("Python SDK: Server")

Advanced Search Patterns

# Find specific implementation patterns
fastmcp_requirements = find_implementation_requirements("FastMCP")

# Search for error handling examples
error_examples = search_content_by_text("try:", max_results=20)

# Locate transport configuration
transport_sections = search_sections_by_name("transport", case_sensitive=False)

Working with Search Results

Each search result includes rich metadata:

• Section name and title
• Document source (MCP Documentation or Python SDK)
• Header level for navigation hierarchy
• Word count for content estimation
• Match context or preview snippets
• Match frequency for relevance ranking

🏗️ Architecture

Document Parser

The DocumentParser class handles:

• Markdown parsing with header-level organization
• Section extraction based on heading hierarchy
• Metadata generation including word counts and document classification
• Search indexing for efficient content retrieval

Search Algorithms

• Name-based search: Pattern matching in section titles and headers
• Content-based search: Full-text search with contextual snippet extraction
• Requirement extraction: Intelligent parsing of code blocks and dependencies
• Relevance ranking: Results sorted by match frequency and section importance

Content Organization

• Hierarchical structure: Maintains document header hierarchy (H1-H6)
• Cross-referencing: Links between related concepts and implementations
• Metadata enrichment: Automatic classification and statistics generation

🔧 Configuration

Customizing Search Behavior

Modify search parameters in the DocumentParser class:

• Adjust context window size for content search
• Configure result ranking algorithms
• Customize section parsing patterns

Adding Content Sources

To add additional documentation sources:

1. Add content to the parser initialization
2. Update document type classification
3. Ensure proper markdown parsing

🐛 Troubleshooting

Common Issues

Server not connecting:

• Ensure MCP Python SDK is properly installed
• Check that the server file has proper permissions
• Verify no port conflicts exist

Search returns no results:

• Check keyword spelling and capitalization
• Try broader search terms
• Use list_all_sections() to see available content

Performance issues:

• Reduce max_results parameter for large searches
• Use more specific keywords to narrow results
• Consider case-sensitive search for exact matches

Debug Mode

Run with verbose logging:

mcp dev mcp_docs_server.py --verbose

🤝 Contributing

Development Setup

1. Clone or download the server file
2. Install development dependencies
3. Run tests with the MCP Inspector

Adding Features

• New search algorithms: Extend the DocumentParser class
• Additional metadata: Enhance section parsing logic
• Custom resources: Add new resource endpoints
• Tool improvements: Extend existing tool functionality

📄 License

This project follows the same license as the MCP Python SDK (MIT License).

🔗 Related Links

• Model Context Protocol Documentation
• MCP Python SDK
• MCP Specification
• Claude Desktop

🆘 Support

For issues related to:

• MCP Documentation Search Server: Create an issue in your repository
• MCP Protocol: Visit MCP GitHub Discussions
• Python SDK: Check the official SDK documentation

---

Made with ❤️ for the MCP Community