claude-context-local

Project Philosophy: This project maintains the full history of the original claude-context for transparency and attribution, but represents a divergent branch focused on privacy and local operation. While the original continues toward cloud integration, we're committed to a 100% local, privacy-first approach. Both projects serve different needs in the ecosystem.

A local-first implementation of claude-context MCP server that replaces cloud dependencies with local alternatives for complete privacy and zero API costs.

🚀 Key Features

• 100% Local: No external API calls, all processing happens on your machine
• PostgreSQL + pgvector: Replaces Milvus/Zilliz Cloud with local PostgreSQL
• Ollama Embeddings: Uses free, open-source models instead of OpenAI
• Privacy-First: Your code never leaves your machine
• Zero API Costs: No subscription fees or usage charges
• Multimodal Support: Nomic embedding model supports text and images

⚡ Quick Start

Prerequisites

1. Install PostgreSQL and pgvector:

brew install postgresql@14 pgvector
brew services start postgresql@14

2. Install and configure Ollama:

brew install ollama
ollama serve  # Start in a separate terminal
ollama pull DC1LEX/nomic-embed-text-v1.5-multimodal

3. Setup database:

psql -U postgres -c "CREATE DATABASE embeddings;"
psql -U postgres -d embeddings -c "CREATE EXTENSION IF NOT EXISTS vector;"

Installation

# Clone the repository
git clone https://github.com/MikeO-AI/claude-context-local.git
cd claude-context-local

# Install dependencies
pnpm install

# Build the project
pnpm build

# Test the integration
node test-postgres.js

Running the MCP Server

Method 1: Direct Installation (Recommended)

# First, validate your setup
node validate-mcp-setup.js

# Start with default settings (PostgreSQL on localhost, Ollama on localhost:11434)
node packages/mcp/dist/index-postgres.js

# Or with custom configuration
POSTGRES_HOST=your-db POSTGRES_PASSWORD=your-pass node packages/mcp/dist/index-postgres.js

Method 2: NPM Package

Install globally from npm:

npm install -g @mikeo-ai/claude-context-local-mcp
claude-context-local-mcp

# Or use with npx (recommended)
npx @mikeo-ai/claude-context-local-mcp@latest

Method 3: Claude MCP Add Command

claude mcp add claude-context-local \
  -e EMBEDDING_PROVIDER=Ollama \
  -e POSTGRES_DATABASE=embeddings \
  -- npx @mikeo-ai/claude-context-local-mcp@latest

For detailed setup instructions, see MCP-SETUP-GUIDE.md

⚙️ Configuration

Claude Code Integration

Add to your Claude Code config (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "claude-context-local": {
      "command": "node",
      "args": ["/absolute/path/to/claude-context-local/packages/mcp/dist/index-postgres.js"],
      "env": {
        "EMBEDDING_PROVIDER": "Ollama",
        "OLLAMA_MODEL": "DC1LEX/nomic-embed-text-v1.5-multimodal",
        "OLLAMA_HOST": "http://localhost:11434",
        "POSTGRES_HOST": "localhost",
        "POSTGRES_PORT": "5432",
        "POSTGRES_DATABASE": "embeddings",
        "POSTGRES_USER": "postgres",
        "POSTGRES_PASSWORD": "postgres"
      }
    }
  }
}

Environment Variables

Variable	Default	Description
EMBEDDING_PROVIDER	Ollama	Embedding provider (Ollama recommended)
OLLAMA_MODEL	DC1LEX/nomic-embed-text-v1.5-multimodal	Ollama model to use
OLLAMA_HOST	http://localhost:11434	Ollama server URL
POSTGRES_HOST	localhost	PostgreSQL host
POSTGRES_PORT	5432	PostgreSQL port
POSTGRES_DATABASE	embeddings	Database name
POSTGRES_USER	postgres	Database user
POSTGRES_PASSWORD	postgres	Database password

🐳 Docker Setup (Optional)

For easier deployment, use Docker Compose:

# docker-compose.yml
version: '3.8'
services:
  postgres:
    image: pgvector/pgvector:pg16
    environment:
      POSTGRES_DB: embeddings
      POSTGRES_USER: postgres
      POSTGRES_PASSWORD: postgres
    ports:
      - "5432:5432"
    volumes:
      - postgres_data:/var/lib/postgresql/data

  ollama:
    image: ollama/ollama
    ports:
      - "11434:11434"
    volumes:
      - ollama_data:/root/.ollama
    command: serve

volumes:
  postgres_data:
  ollama_data:

Then:

docker-compose up -d
docker exec -it claude-context-local_ollama_1 ollama pull DC1LEX/nomic-embed-text-v1.5-multimodal

🔧 MCP Tools

add_codebase

Index a codebase directory for semantic search:

add_codebase({
  path: "/path/to/your/project"
})

search_codebase

Search through indexed codebases:

search_codebase({
  query: "authentication logic",
  path: "/path/to/your/project",
  limit: 10
})

📊 Comparison with Original

Feature	Original (Milvus)	Local (PostgreSQL)
Vector Database	Milvus/Zilliz Cloud	PostgreSQL + pgvector
Embeddings	OpenAI/VoyageAI	Ollama (local)
Privacy	Data sent to cloud	100% local
Cost	API usage fees	Free
Setup	Cloud account required	Local installation
Internet	Required	Not required
Dimensions	Variable	768 (configurable)

🙏 Credits & Attribution

This project is based on claude-context by Zilliz, licensed under the MIT License.

What's Changed

• Replaced Milvus/Zilliz Cloud with PostgreSQL + pgvector
• Replaced OpenAI embeddings with Ollama
• Added multimodal embedding support
• Removed all external API dependencies
• Added comprehensive PostgreSQL implementation

Original Contributors

• Zilliz Team - Original claude-context implementation
• Cheney Zhang - Original author

📚 Documentation

• PostgreSQL Setup Guide - Detailed PostgreSQL configuration
• Test Script - Verify your installation
• Original Documentation - Reference the original project

🔍 Troubleshooting

Common Issues

1. pgvector not found

   brew install pgvector
   psql -U postgres -d embeddings -c "CREATE EXTENSION vector;"

2. Ollama connection failed

   ollama serve  # Start Ollama service

3. Model not found

   ollama pull DC1LEX/nomic-embed-text-v1.5-multimodal

🗺️ Roadmap

• SQLite support for simpler setup
• Additional Ollama model support
• Web UI for codebase management
• Incremental indexing
• Docker one-click deployment
• Support for more local embedding models

🤝 Contributing

Contributions are welcome! This is a community project aimed at providing a fully local alternative to cloud-based code context solutions.

How to Contribute

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

Areas for Contribution

• Support for additional databases (SQLite, DuckDB)
• More embedding model integrations
• Performance optimizations
• Documentation improvements
• Testing and bug fixes

📄 License

MIT License - see LICENSE for details.

This project includes code from claude-context by Zilliz, also licensed under MIT.

⭐ Star History

If you find this project useful, please consider giving it a star!

---

Built with ❤️ for the community by MikeO-AI