Chroma MCP Server

🚀 Chroma Project

Chroma is a tool designed to create persistent and searchable "working memory" to assist in AI development workflows. The server includes specialized tools that leverage embedding technology to manage context across sessions.

🚀 Quick Start

📦 Installation

Install with Hatch

To install the Chroma project using the hatch tool, run the following commands:

pip install hatch
cd path/to/chroma-project
hatch install --dev

For more information, refer to the Hatch documentation.

Install with Uv

To install the Chroma project from PyPI using the uv tool, run the following command:

uvx install chroma-mcp-server

For more information, refer to the Uv documentation.

💻 Usage Examples

Run the Server

To run the Chroma server, use the following command:

hatch run chroma-mcp-server start -- --config path/to/config.yaml

Or run it directly from the version installed from PyPI:

chroma-mcp-server start -- --config path/to/config.yaml

For more information, refer to the Quick Start Guide.

Integration Guide

Integration with Cursor

To integrate Chroma into your project, create or modify the .cursor/mcp.json file in the project root directory. Here is a configuration example:

{
  "mcpServers": {
    "chroma_dev": {
      "command": "/path/to/project/scripts/run_chroma_mcp_server_dev.sh",
      "args": [],
      "env": {
        "CHROMA_CLIENT_TYPE": "persistent",
        "CHROMA_DATA_DIR": "/path/to/your/dev_data",
        "CHROMA_LOG_DIR": "/path/to/your/dev_logs",
        "LOG_LEVEL": "DEBUG",
        "MCP_LOG_LEVEL": "DEBUG"
      }
    },
    "chroma_test": {
      "command": "uvx",
      "args": [
        "--refresh",
        "--default-index", "https://test.pypi.org/simple/",
        "--index", "https://pypi.org/simple/",
        "--index-strategy", "unsafe-best-match",
        "chroma-mcp-server@latest"
      ],
      "env": {
        "CHROMA_CLIENT_TYPE": "persistent",
        "CHROMA_DATA_DIR": "/path/to/your/test_data",
        "CHROMA_LOG_DIR": "/path/to/your/test_logs",
        "LOG_LEVEL": "INFO",
        "MCP_LOG_LEVEL": "INFO"
      }
    },
    "chroma_prod": {
      "command": "uvx",
      "args": [
        "chroma-mcp-server"
      ],
      "env": {
        "CHROMA_CLIENT_TYPE": "persistent",
        "CHROMA_DATA_DIR": "/path/to/your/prod_data",
        "CHROMA_LOG_DIR": "/path/to/your/prod_logs",
        "LOG_LEVEL": "INFO",
        "MCP_LOG_LEVEL": "INFO"
      }
    }
  }
}

For more information, refer to the Integration Guide.

Development Process

To contribute to the development, run the following commands:

git clone git@github.com:your-username/chroma-project.git
cd chroma-project
pip install -r requirements/development.txt
hatch run pytest tests/

For more information, refer to the Developer Documentation.

🔧 Technical Details

Motivation

To address the challenges in AI development workflows, we developed Chroma. By combining modern storage technologies with distributed system design, Chroma can efficiently manage large-scale data.

Features

• Persistent Storage: Ensures data remains available across sessions.
• Scalability: Supports high-concurrency access and large data volume processing.
• Integration Capability: Compatible with the workflows of various tools and services.

For more information, refer to the Design Documentation.

📚 Documentation

• Project Repository: https://github.com/your-username/chroma-project
• Online Documentation: https://your-docs-site.chroma-project.org

🤝 Contribution Guide

Community contributions are welcome! Please refer to the Developer Documentation to learn how to contribute.

⚠️ Important Note

This document is an example, and actual projects should be adjusted according to specific requirements.