CrackingShells/Hatch-MCP-Server

Hatch-MCP-Server

A wrapper around FastMCP servers that adds automatic citation capabilities for scientific transparency and attribution in MCP (Model Context Protocol) applications.

What is HatchMCP?

HatchMCP enhances FastMCP servers by automatically exposing citation information as standardized MCP resources. This enables proper attribution for both original research/algorithms and MCP server implementations, which is critical for attribution in scientific computing and autonomous agent systems.

Key Features:

• 🔬 Scientific Attribution: Automatic citation resource registration
• 🔄 FastMCP Compatibility: Works with existing FastMCP servers
• 📋 Standardized URIs: Uses citation://origin/ and citation://mcp/ schemes
• 🪶 Lightweight: Minimal overhead, zero configuration required

Quick Start

Installation

# Install directly from the repository
pip install git+https://github.com/CrackingShells/Hatch-MCP-Server.git

# Or install local copy
git clone https://github.com/CrackingShells/Hatch-MCP-Server.git
cd Hatch-MCP-Server
pip install .

Basic Usage

from hatch_mcp_server import HatchMCP

# Create an MCP server with citation capabilities
hatch_mcp = HatchMCP(
    name="MyServer",
    origin_citation="Smith, J. 'Original Algorithm', Nature, 2024",
    mcp_citation="Your Name, 'MCP Implementation', GitHub, 2024"
)

# Add tools using FastMCP decorator syntax
@hatch_mcp.server.tool()
def process_data(data: str) -> str:
    """Process data using the wrapped algorithm."""
    hatch_mcp.logger.info(f"Processing: {data}")
    return f"Processed: {data}"

# Run the server
if __name__ == "__main__":
    hatch_mcp.server.run()

Existing FastMCP Server

Assuming you have an existing FastMCP server (mcp_server.py)

from mcp.server.fastmcp import FastMCP

existing_server = FastMCP("my_existing_server")

@existing_server.tool()
def existing_function(data: str) -> str:
    return f"Existing: {data}"

You can wrap it with HatchMCP:

from mcp.server.fastmcp import FastMCP
from hatch_mcp_server import HatchMCP

from mcp_server import existing_server

# Wrap with HatchMCP for citation capabilities
hatch_mcp = HatchMCP(
    name="my_existing_server",
    fast_mcp=existing_server,
    origin_citation="Original work citation",
    mcp_citation="MCP implementation citation"
)

# Run the server
if __name__ == "__main__":
    hatch_mcp.server.run()

The server automatically exposes citation information through:

• citation://origin/MyServer - Original algorithm citation
• citation://mcp/MyServer - MCP implementation citation

Documentation

📚 - Full documentation table of contents and navigation guide

Quick Start Guides

• - Installation, setup, and first steps
• - Complete API documentation
• - Understanding scientific attribution features
• - Practical usage patterns and code samples

Developer Resources

• - Complete development workflow and MCP-specific requirements
• - Design principles and implementation details

I Want To

• 🚀 Get started quickly →
• 🔄 Wrap existing FastMCP servers →
• 📖 Learn about citations →
• 🔍 Look up specific APIs →
• 💡 See examples →
• 🛠️ Contribute code →
• 🏗️ Understand design →
• 📋 Browse all docs →

Why Use HatchMCP?

Scientific Transparency

# Citation information is automatically available to MCP clients
# citation://origin/ServerName -> "Smith, J. 'Algorithm X', Nature, 2024"
# citation://mcp/ServerName -> "Your Name, 'MCP Implementation', 2024"

Zero Configuration

# Just wrap your server - citation resources are registered automatically
hatch_mcp = HatchMCP("server_name", origin_citation="...", mcp_citation="...")
# Use hatch_mcp.server exactly like FastMCP

FastMCP Compatibility

# Wrap existing FastMCP servers without changes
existing_server = FastMCP("my_server")
wrapped = HatchMCP("my_server", fast_mcp=existing_server, ...)

Contributing

We welcome contributions! Please see our comprehensive for detailed guidelines covering development workflow, MCP compatibility requirements, and coding standards.

Overview

1. Fork and clone the repository
2. Install dependencies: pip install -e . and npm install
3. Create a feature branch: git checkout -b feat/your-feature
4. Make changes and add tests
5. Use conventional commits: npm run commit for guided commits
6. Create a pull request

For complete details on commit formats, testing requirements, and MCP-specific development guidelines, see the .

Requirements

• Python: 3.12+
• Dependencies: mcp>=1.6.0
• Compatibility: FastMCP servers and MCP clients

License

AGPL v3: see

Part of the Cracking Shells Ecosystem: Hatch! - A package management system for scientific MCP servers.