Skip to main content

Meta MCP Server for MCP development intelligence

Project description

mcp4mcp

Meta MCP Server - AI-powered development assistant for building better MCP projects

Python 3.8+ FastMCP License: Unlicense

mcp4mcp automatically tracks your MCP tools, detects duplicates, suggests improvements, and provides AI-powered guidance throughout your development process.

โšก Zero-Install Setup (Recommended)

Like npx for Node.js, use uvx for instant Python package execution:

{
  "mcpServers": {
    "mcp4mcp": {
      "command": "uvx",
      "args": ["--from", "mcp4mcp", "mcp4mcp-server"]
    }
  }
}

No installation needed! Claude Desktop automatically downloads and runs mcp4mcp when needed.

๐Ÿš€ Quick Start

Installation

# Install mcp4mcp from PyPI
pip install mcp4mcp

Setup with Claude Desktop

Add this to your Claude Desktop claude_desktop_config.json file:

{
  "mcpServers": {
    "mcp4mcp": {
      "command": "mcp4mcp-server",
      "args": []
    }
  }
}

Claude Desktop will automatically start and manage the mcp4mcp server. Then just chat:

"Help me build an MCP server for file processing. Use mcp4mcp to track progress and avoid duplicates."

Manual Usage (Optional)

# Or run manually for testing
mcp4mcp-server

# Or run the demo
mcp4mcp demo

Using with Claude Desktop

mcp4mcp integrates seamlessly with Claude Desktop for AI-powered MCP development:

# 1. Start mcp4mcp server (in one terminal)
mcp4mcp-server

# 2. Configure Claude Desktop MCP settings
# Add to your Claude Desktop configuration:
# Server Name: mcp4mcp
# Command: mcp4mcp-server

# 3. Chat with Claude Desktop about MCP development
"Help me build an MCP server for database operations. Use mcp4mcp to track progress."

What Claude Desktop can do with mcp4mcp:

  • ๐ŸŽฏ Track development sessions automatically
  • ๐Ÿ” Check for duplicate tools before building
  • ๐Ÿ’ก Provide AI suggestions based on your project state
  • ๐Ÿ“Š Monitor progress across development sessions
  • ๐Ÿ”„ Discover existing tools in your codebase

Development Installation

# Install from source for development
git clone https://github.com/hmatt1/mcp4mcp.git
cd mcp4mcp && pip install -e .
python server.py

๐Ÿ’ก What You Get

  • ๐Ÿง  AI suggestions for next development steps
  • ๐Ÿ” Duplicate detection before you build conflicting tools
  • ๐Ÿ“Š Progress tracking across development sessions
  • ๐Ÿ”„ Auto-discovery of tools in your codebase
  • ๐Ÿ“ˆ Analytics on your development patterns

๐Ÿ› ๏ธ Core Tools

Tool Purpose
get_project_state_tool Load your current project and tools
update_project_state_tool Add/update tools and project info
scan_project_files_tool Auto-discover tools in your code
check_before_build_tool Check for conflicts before building
suggest_next_action_tool Get AI-powered development suggestions
analyze_tool_similarity_tool Find similar/duplicate tools
track_development_session_tool Log your development activities
get_session_analytics_tool View development insights

๐Ÿค– Using with Claude Desktop

mcp4mcp is designed to work perfectly with Claude Desktop for intelligent MCP development:

Quick Setup

# Terminal 1: Start mcp4mcp server
mcp4mcp-server

# Terminal 2: Configure Claude Desktop MCP settings
# In Claude Desktop:
# - Go to Settings โ†’ MCP Servers
# - Add new server:
#   Name: mcp4mcp
#   Command: mcp4mcp-server
#   Args: (leave empty)

Example Claude Desktop Sessions

Building a New MCP Server

You: "Help me build an MCP server for file processing with CSV and JSON support."

Claude Desktop automatically:
1. ๐ŸŽฏ Calls track_development_session_tool("file-processing-mcp", "Building CSV/JSON tools")
2. ๐Ÿ” Uses check_before_build_tool to see if similar tools exist
3. ๐Ÿ’ก Calls suggest_next_action_tool for personalized guidance
4. ๐Ÿ“Š Tracks progress as you build each tool

Avoiding Duplicate Work

You: "I want to add a data validation tool to my project."

Claude Desktop automatically:
1. ๐Ÿ” Scans existing tools with scan_project_files_tool
2. โš ๏ธ  Alerts if similar validation tools already exist
3. ๐Ÿ’ก Suggests reusing or extending existing tools instead
4. ๐ŸŽฏ Tracks the decision in your development session

Getting Development Insights

You: "What should I work on next for my MCP project?"

Claude Desktop automatically:
1. ๐Ÿ“Š Calls get_session_analytics_tool for project insights
2. ๐Ÿ’ก Uses suggest_next_action_tool based on current state
3. ๐ŸŽฏ Provides personalized recommendations
4. ๐Ÿ“ˆ Shows development patterns and progress

Benefits of Claude Desktop + mcp4mcp

  • ๐Ÿง  Intelligent Context: Claude Desktop understands MCP development patterns
  • ๐Ÿ” Automatic Conflict Detection: Prevents duplicate tool development
  • ๐Ÿ“Š Continuous Tracking: Every development action is logged automatically
  • ๐Ÿ’ก Contextual Suggestions: AI guidance based on your specific project state
  • ๐Ÿš€ Accelerated Development: Focus on building, not project management

๐Ÿ“‹ Usage Examples

Start Development Session

# Log what you're working on
await track_development_session(
    "Building file processing tools", 
    "my_project",
    "csv_reader"
)

Check Before Building

# Avoid duplicates
result = await check_before_build(
    "file_processor", 
    "Process CSV files",
    "my_project" 
)

if result['conflicts']:
    print("โš ๏ธ Similar tools exist - consider reusing instead")

Get AI Suggestions

# Get personalized guidance
suggestions = await suggest_next_action(
    "my_project",
    "Just finished the CSV reader, what's next?"
)

for suggestion in suggestions['suggestions']:
    print(f"๐Ÿ’ก {suggestion}")

Auto-Discover Tools

# Scan your codebase
result = await scan_project_files("my_project", "./src")
print(f"๐Ÿ” Found {result['new_tools']} new tools")

๐Ÿค– AI-Powered MCP Development

Use this prompt template with any LLM to build MCP servers that leverage mcp4mcp:

# MCP Server Development with mcp4mcp

You are an expert MCP (Model Context Protocol) developer building a new MCP server. 
You have access to mcp4mcp tools that provide intelligent development assistance.

## Your Development Process:

1. **Start Each Session**: Always begin by calling `track_development_session_tool` 
   to log what you're working on

2. **Before Building Any Tool**: Call `check_before_build_tool` to check for 
   conflicts and similar existing tools

3. **Get AI Guidance**: Use `suggest_next_action_tool` for personalized 
   development recommendations based on project state

4. **Update Progress**: Use `update_project_state_tool` to track tools as you 
   build them (planned โ†’ in_progress โ†’ completed)

5. **Discover Existing Tools**: Use `scan_project_files_tool` to automatically 
   find tools in the codebase

6. **Check for Duplicates**: Run `analyze_tool_similarity_tool` periodically 
   to find similar tools that could be consolidated

## Current Task:
Build a [DOMAIN] MCP server with tools for [SPECIFIC_FUNCTIONALITY].

## Project Details:
- Project name: [PROJECT_NAME]
- Description: [PROJECT_DESCRIPTION] 
- Key requirements: [LIST_REQUIREMENTS]

Start by calling the appropriate mcp4mcp tools to understand the current state 
and get AI-powered suggestions for the best approach.

Example Prompt Usage:

# MCP Server Development with mcp4mcp

You are an expert MCP developer building a new MCP server. You have access to mcp4mcp tools.

## Current Task:
Build a file processing MCP server with tools for reading, writing, and transforming CSV/JSON files.

## Project Details:
- Project name: file-processor-mcp
- Description: MCP server for file operations with data transformation capabilities
- Key requirements: 
  * Read CSV and JSON files
  * Write data in multiple formats  
  * Transform data between formats
  * Validate file schemas
  * Handle large files efficiently

Start by calling mcp4mcp tools to check current state and get development guidance.

๐Ÿ”ง Integration

Add mcp4mcp to any FastMCP project:

from fastmcp import FastMCP
from mcp4mcp.tools.state_management import register_state_tools
from mcp4mcp.tools.intelligence import register_intelligence_tools  
from mcp4mcp.tools.tracking import register_tracking_tools

# Your MCP server
mcp = FastMCP("your-server")

# Add mcp4mcp intelligence
register_state_tools(mcp)
register_intelligence_tools(mcp) 
register_tracking_tools(mcp)

# Your tools
@mcp.tool()
def your_tool():
    return "Hello World"

mcp.run()

๐Ÿ“Š Development Analytics

View your development patterns:

# Get insights on your development
analytics = await get_session_analytics("my_project", days=7)

print(f"๐Ÿ“ˆ This week:")
print(f"  Sessions: {analytics['total_sessions']}")  
print(f"  Time: {analytics['total_development_time']}")
print(f"  Tools: {len(analytics['tools_worked_on'])}")

๐Ÿ—ƒ๏ธ Data Storage

All data stored locally in ~/.mcp4mcp/projects.db - no external dependencies.

๐Ÿงช Testing

# Run demo
python main.py demo

# Run tests  
python main.py test

# FastMCP diagnostics
python run_diagnostic.py

๐Ÿ› ๏ธ Development

# Setup
git clone https://github.com/hmatt1/mcp4mcp.git
cd mcp4mcp
pip install -e ".[dev]"

# Test
python -m pytest tests/ -v

๐Ÿ“„ License

This is free and unencumbered software released into the public domain. See the UNLICENSE file for details.

๐Ÿค Support


mcp4mcp - Intelligence for MCP development ๐Ÿง โœจ

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

mcp4mcp-0.1.32.tar.gz (66.6 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

mcp4mcp-0.1.32-py3-none-any.whl (24.4 kB view details)

Uploaded Python 3

File details

Details for the file mcp4mcp-0.1.32.tar.gz.

File metadata

  • Download URL: mcp4mcp-0.1.32.tar.gz
  • Upload date:
  • Size: 66.6 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.1.0 CPython/3.12.3

File hashes

Hashes for mcp4mcp-0.1.32.tar.gz
Algorithm Hash digest
SHA256 1c95f4e50fc4e0534a6681fccd8808282bec99f9768e28527151a20a17601257
MD5 a8d1573ffc29b3836084329ee6ac0f09
BLAKE2b-256 4b50a61babe628fdb4ad9c3c69be7904606e7c97a953bbcb8e845cdb3339f7c6

See more details on using hashes here.

File details

Details for the file mcp4mcp-0.1.32-py3-none-any.whl.

File metadata

  • Download URL: mcp4mcp-0.1.32-py3-none-any.whl
  • Upload date:
  • Size: 24.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.1.0 CPython/3.12.3

File hashes

Hashes for mcp4mcp-0.1.32-py3-none-any.whl
Algorithm Hash digest
SHA256 e0987a122c448e52355efa9311298bfd7c3336b26670d98b49ecc128afaf53c8
MD5 0585b02d960b6b2d4722f9bb496a3329
BLAKE2b-256 b7484402968c693b67bdad94f2ff726e1219f0edf193ed218a1786e37b8c5360

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page