Supports configuration through .env files for storing API keys and other environment variables needed for the MCP server.
Utilizes Google Gemini AI to perform project planning, code reviews, and execution analysis, acting as an AI architect that provides structured project plans, code quality assessment, security vulnerability detection, and debugging assistance.
Leverages Pydantic for data validation and settings management in the MCP server's internal models.
Uses Rich for enhanced terminal output and formatting when displaying MCP server information and logs.
Plan-MCP
A Model Context Protocol (MCP) server that leverages Google Gemini AI for intelligent project planning and code review.
🌟 Overview
Plan-MCP acts as an AI-powered project architect that bridges Gemini's planning capabilities with Claude's coding abilities:
Gemini as Architect: Analyzes requirements, creates project plans, reviews code quality
Claude as Developer: Implements code based on Gemini's guidance
Continuous Feedback Loop: Gemini reviews execution results and provides iterative improvements
🚀 Features
Plan-MCP provides complete MCP feature support, making it one of the most comprehensive MCP servers available:
✅ Complete MCP Feature Matrix
Feature | Status | Description |
Resources | ✅ | File system access (file://, dir://, workspace://) |
Prompts | ✅ | 4 structured prompt templates for common tasks |
Tools | ✅ | 10 comprehensive tools for project management |
Discovery | ✅ | Dynamic tool discovery (handled by FastMCP) |
Sampling | ✅ | LLM text generation for documentation and tests |
Roots | ✅ | Workspace navigation and project root suggestions |
Elicitation | ✅ | Interactive user input collection |
🔧 Core Tools
1. Project Planning (plan_project
)
Break down complex requirements into structured phases and tasks
Generate detailed project plans with priorities and dependencies
Estimate effort and identify potential risks
Support for technical constraints and preferred tech stacks
2. Code Review (review_code
)
Comprehensive code quality analysis
Security vulnerability detection
Performance optimization suggestions
Best practices and design pattern recommendations
Language-agnostic support
3. Execution Analysis (analyze_execution
)
Debug runtime errors with root cause analysis
Provide specific code fixes with explanations
Evaluate if execution meets expected behavior
Guide iterative development with next steps
4. Directory Review (review_directory
)
Complete project/directory analysis
Multi-file code quality assessment
Project structure recommendations
Security scanning across entire codebase
🎯 Advanced Features
Interactive Tools (Elicitation)
Interactive Project Planning: Collects user preferences and requirements dynamically
Interactive Code Review: Customizes review focus based on user needs
LLM Sampling
Documentation Generation: Auto-generates comprehensive docs for code
Test Generation: Creates unit tests with proper assertions and edge cases
File System Resources
File Access: Read individual files with
file://
URIsDirectory Access: Access entire directories with
dir://
URIsWorkspace Navigation: Current workspace info with
workspace://current
Workspace Management (Roots)
Workspace Roots: Lists available workspace directories
Project Suggestions: Recommends appropriate project locations by type
Prompt Templates
Code Review Template: Structured code review prompts
Project Planning Template: Interactive planning conversations
Debug Assistant: Systematic debugging guidance
Architecture Review: System architecture analysis
📋 Prerequisites
Python 3.10 or higher
Google Gemini API key
Claude Code (for MCP integration)
🛠️ Installation
Quick Start with uvx (Recommended)
Traditional pip Installation
🔧 Configuration
Set up your Gemini API key
Or create a .env
file:
Claude Code Integration
🚀 Method 1: Direct from GitHub (Recommended)
Run directly from GitHub using uv
without local installation:
This creates a .mcp.json
file in your project root. For secure API key management, edit the file:
🔧 Method 2: Local Installation (Recommended)
Install locally for reliable connection:
✅ Verify Installation
Check if the MCP server is working:
Alternative Configuration Options
Personal global configuration:
Local testing configuration:
Managing MCP Services
💻 Usage
Once configured, you can use these tools in Claude Code:
1. Create a project plan
2. Review code
3. Review entire directory/project
4. Analyze execution errors
5. Access files and directories
🏗️ Architecture
🤝 Workflow Example
Human → Claude: "Help me build a web scraper"
Claude → Plan-MCP: Requests project plan
Plan-MCP → Gemini: Analyzes requirements
Gemini → Plan-MCP: Returns structured plan
Plan-MCP → Claude: Delivers plan
Claude: Implements first task
Claude → Plan-MCP: Submits code for review
Plan-MCP → Gemini: Reviews code
Gemini → Plan-MCP: Provides feedback
Plan-MCP → Claude: Delivers improvements
Cycle continues...
📚 API Reference
Tools
plan_project
Description: Create a comprehensive project plan
Parameters:
description
(required): Project descriptionrequirements
: List of specific requirementsconstraints
: Project constraintstech_stack
: Preferred technologies
review_code
Description: Review code for quality and issues
Parameters:
code
(required): Code to reviewlanguage
(required): Programming languagecontext
: Additional contextfocus_areas
: Specific areas to focus on
analyze_execution
Description: Analyze execution results and debug errors
Parameters:
code
(required): Code that was executedexecution_output
(required): Output or error messagesexpected_behavior
: What the code should doerror_messages
: Specific error messageslanguage
: Programming language (default: python)
🧪 Development
Set up development environment
Code quality
🐛 Troubleshooting
Common Issues
"GEMINI_API_KEY not found"
Ensure your API key is set in environment variables:
export GEMINI_API_KEY="your_key_here"
Or create a
.env
file in your working directory withGEMINI_API_KEY=your_key_here
Get your API key from: https://makersuite.google.com/app/apikey
Connection errors
Verify your internet connection
Check if the Gemini API is accessible
Ensure your API key has proper permissions
MCP connection issues
Restart Claude Code after configuration
Check that the server starts without errors
Look at Claude Code logs for errors
📄 License
MIT License - see LICENSE file for details
🙏 Acknowledgments
Google Gemini for powerful AI capabilities
Anthropic for Claude and the MCP protocol
The open-source community for inspiration
This server cannot be installed
An MCP server that uses Google Gemini AI to analyze requirements, create project plans, review code quality, and provide execution analysis feedback for software development projects.
Related MCP Servers
- AsecurityFlicenseAqualityAn MCP server implementation that leverages Google's Gemini API to provide analytical problem-solving capabilities through sequential thinking steps without code generation.Last updated -20
- AsecurityFlicenseAqualityA server that provides access to Google Gemini AI capabilities including text generation, image analysis, YouTube video analysis, and web search functionality through the MCP protocol.Last updated -644
- AsecurityAlicenseAqualityA dedicated server that wraps Google's Gemini AI models in a Model Context Protocol (MCP) interface, allowing other LLMs and MCP-compatible systems to access Gemini's capabilities like content generation, function calling, chat, and file handling through standardized tools.Last updated -1632MIT License
- -securityAlicense-qualityAn MCP server that enables other AI models (like Claude) to use Google's Gemini models as tools for specific tasks through a standardized interface.