MCP Titan

# Titan Memory MCP Server Documentation Welcome to the Titan Memory MCP Server documentation! This documentation will help you understand, install, and use the server effectively. ## Table of Contents ### Getting Started - [Installation](../README.md#installation) - [Quick Start](../README.md#usage) - [Features](../README.md#features) ### Guides - [Cursor IDE Integration](guides/cursor.md) - [Basic Usage Examples](examples/basic-usage.md) ### API Reference - [Server API](api/README.md) - [HTTP Endpoints](api/README.md#http-api) - [Tool Reference](api/README.md#available-tools) ### Development - [Contributing Guidelines](../CONTRIBUTING.md) - [Development Setup](../CONTRIBUTING.md#development-setup) - [Testing](../CONTRIBUTING.md#testing) ### Additional Resources - [GitHub Repository](https://github.com/henryhawke/mcp-titan) - [Issue Tracker](https://github.com/henryhawke/mcp-titan/issues) - [Model Context Protocol](https://modelcontextprotocol.io) ## Overview The Titan Memory MCP Server is a Model Context Protocol implementation that provides memory-augmented learning capabilities for Cursor IDE. It maintains a persistent memory state that evolves based on interactions, enabling contextual awareness and learning over time. ### Key Features - 🧠 Automatic memory management - 🔄 Real-time updates - 📊 Memory state analysis - 🔌 Cursor IDE integration - 🚀 HTTP API support - 💾 Persistent storage ### Architecture ```mermaid graph TD A[Cursor IDE] -->|MCP| B[Titan Memory Server] B --> C[Memory State] B --> D[TensorFlow.js] B --> E[HTTP API] C --> F[Persistent Storage] ``` The server automatically saves the memory state every 5 minutes, and it's initialized with zeros when first created. The memory vectors are 768-dimensional by default, which is a common size for language model embeddings. ### Getting Help 1. Check the [documentation](api/README.md) 2. Look through [examples](examples/basic-usage.md) 3. Search [existing issues](https://github.com/henryhawke/mcp-titan/issues) 4. Create a new issue if needed ### Contributing We welcome contributions! Please see our [Contributing Guidelines](../CONTRIBUTING.md) for details on: - Code style - Pull request process - Development setup - Testing requirements ### License This project is licensed under the MIT License - see the [LICENSE](../LICENSE) file for details.