<div align="center">
  <a href="https://matterai.so">
    <img
      src="https://matterai.so/favicon.png"
      alt="Matter AI Logo"
      height="64"
    />
  </a>
  <br />
  <p>
    <h3>
      <b>
        MatterAI
      </b>
    </h3>
  </p>
  <p>
    <b>
      AI Super Intelligence for Engineering
    </b>
  </p>
  <p>

![Matter Og Image](https://res.cloudinary.com/dxvbskvxm/image/upload/v1756191766/og-image-matter_s5m5l6.webp)

  </p>
</div>


# Matter AI MCP Server

MatterAI MCP offers code reviews right in your IDE when using AI Agents such as in Cursor, Windsurf, VS Code, Cline and more to enhances your development workflow. Built with [FastMCP](https://github.com/modelcontextprotocol/fastmcp) in Python, it provides advanced code review capabilities, implementation planning, and pull request generation to help you release code with confidence.

## Features
- **Code review tools** - Get comprehensive code reviews for individual files or full git diffs
- **Implementation planning** - Generate detailed implementation plans for AI agents
- **Pull request generation** - Create pull requests with auto-generated titles and descriptions
- **Random cat facts** - Because who doesn't love cat facts?

## Requirements
- Python 3.11+
- See `requirements.txt` for dependencies

## Installation
```bash
pip install -r requirements.txt
```

## Setup

### API Key
To use Matter AI MCP Server, you need an API key:
1. Obtain your API key from [https://app.matterai.dev/settings](https://app.matterai.dev/settings)
2. Use this key in your MCP configuration as shown below

### MCP Configuration
Create an MCP configuration file with the following content:

```json
{
  "mcpServers": {
    "matter-ai": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.matterai.so/sse",
        "--header",
        "X-AUTH-TOKEN:MATTER_AI_API_KEY"
      ]
    }
  }
}
```

Replace `MATTER_AI_API_KEY` with your actual API key.

## Usage
Run the server:
```bash
python server.py
```
The server will start on `http://localhost:9000` (default for FastMCP).

### Connecting from Cursor or Windsurf
- Use the MCP (Model Context Protocol) integration
- Point to: `http://localhost:9000/sse`
- Tools will auto-discover and appear in the client

## Tools

### 1. Code Review
`codereview(generated_code: str, git_owner: str, git_repo: str, git_branch: str, git_user: str, languages: str) -> str`
Provides code review for the generated code.

### 2. Full Code Review
`codereview_full(git_diff: str, git_owner: str, git_repo: str, git_branch: str, git_user: str) -> str`
Provides a comprehensive code review based on git diff output.

### 1. Cat Fact
`cat_fact() -> str`
Returns a random cat fact.

## Docker Build and Use

### Building the Docker Image

```bash
docker build -t matter-ai-mcp .
```

### Running the Docker Container

```bash
docker run -p 9000:9000 -e MATTER_API_ENDPOINT=https://api.matterai.so
```

The server will be accessible at `http://localhost:9000`.

## License
MIT

## Resources:
- Website: https://matterai.so
- Docs: https://docs.matterai.so