🐅 Tiger Memory

Claude Code that actually remembers your project

Tiger Memory is a persistent memory system for Claude Code that provides architectural intelligence. It solves the context loss problem by automatically capturing and retrieving project decisions between sessions, evolving into a social architectural intelligence platform where developers learn from patterns across similar projects.

✨ Features

• 🧠 Automatic Decision Capture - Claude automatically stores architectural decisions and reasoning
• 🔍 Intelligent Context Retrieval - Semantic search through project history using vector embeddings
• 🏛️ Architectural Intelligence - Discover patterns from similar projects in the community
• 📊 Project Timeline - Track architectural evolution over time
• ⚡ Zero Manual Work - Fully automatic integration with Claude Code via MCP
• 🌐 Cloud-Native - Remote MCP server with SSE transport for seamless Claude Code integration
• 👥 Multi-User Support - GitHub OAuth authentication with team collaboration
• 🔒 Privacy-First - You control what decisions are shared publicly

🚀 Quick Start

Installation

npm install -g tigermemory

Setup

1. Authenticate and register:

   tigermemory login

   This opens your browser for GitHub OAuth authentication and automatically registers Tiger Memory with Claude Code globally.

That's it! Tiger Memory now works in any directory and will automatically capture and recall your project decisions.

🎯 The Magic Moment

Before Tiger Memory:

You: "Continue with our task app"
Claude: "I'd be happy to help, but I don't have context about your task app. 
         Could you tell me about your current setup, tech stack, and what 
         you're working on?"

After Tiger Memory:

You: "Continue with our task app"
Claude: "I can see we're building a Node.js task management app with Express, 
         PostgreSQL, and JWT authentication. Based on our previous decisions, 
         you wanted to implement user roles with a clean database abstraction layer. 
         Should we continue with the task CRUD endpoints we discussed?"

🛠️ How It Works

Tiger Memory operates through four core MCP tools that Claude Code uses automatically:

1. remember_decision

Automatically called when Claude makes architectural choices:

{
  decision: "Use Express.js for the web framework",
  reasoning: "Express provides flexibility and has excellent middleware ecosystem",
  type: "tech_stack",
  alternatives_considered: ["Fastify", "Koa", "NestJS"],
  files_affected: ["package.json", "src/server.js"],
  confidence: 0.9,
  public: true
}

2. recall_context

Retrieves project context with semantic search:

// Get all recent decisions
await recall_context({ limit: 10 })

// Semantic search
await recall_context({ 
  query: "authentication decisions",
  limit: 5 
})

3. discover_patterns

Find architectural patterns from similar projects:

await discover_patterns({
  query: "user authentication patterns",
  tech_stack: ["nodejs", "express"],
  project_type: "backend"
})

4. get_timeline

View chronological decision history:

// All decisions
await get_timeline({})

// Filtered by date and category
await get_timeline({
  since: "2024-01-01",
  category: "architecture"
})

📋 Supported Project Types

Tiger Memory automatically detects projects using:

• Node.js - package.json
• Python - pyproject.toml, requirements.txt, setup.py
• Rust - Cargo.toml
• Go - go.mod
• Java - pom.xml, build.gradle
• PHP - composer.json
• Ruby - Gemfile
• Elixir - mix.exs
• Deno - deno.json
• Git repositories - .git directory

🗄️ Architecture

Tiger Memory uses a cloud-native remote MCP architecture:

┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐
│   Claude Code   │───▶│  Tiger Memory    │───▶│  Tiger Cloud    │
│                 │    │  Remote MCP      │    │  PostgreSQL     │
│  - remember     │    │  Server (SSE)    │    │  + TimescaleDB  │
│  - recall       │    │                  │    │  + pgvector     │
│  - discover     │    │  - GitHub OAuth  │    │                 │
│  - timeline     │    │  - User Context  │    │  - Users        │
│                 │    │  - Vector        │    │  - Teams        │
│                 │    │    Embeddings    │    │  - Projects     │
│                 │    │  - Intelligence  │    │  - Decisions    │
└─────────────────┘    └──────────────────┘    └─────────────────┘

📖 CLI Commands

tigermemory login

Authenticate with GitHub and register Tiger Memory globally with Claude Code.

tigermemory login               # Login to hosted service and auto-register
tigermemory login --local       # Login to local development instance
tigermemory login --no-register # Skip automatic Claude Code registration

tigermemory status

Check Tiger Memory authentication and configuration status.

tigermemory status

tigermemory server

Start the MCP server (used internally by Claude Code).

tigermemory server

⚙️ Configuration

Environment Variables

| Variable | Required | Description | |----------|----------|-------------| | ANTHROPIC_API_KEY | Yes | Anthropic API key for vector embeddings (development only) | | LOG_LEVEL | No | Logging level (debug, info, warn, error) | | NODE_ENV | No | Environment (development, production) |

Note: For the hosted service, authentication is handled via GitHub OAuth. Environment variables are only needed for local development.

MCP Configuration

Tiger Memory registers globally with Claude Code, so no per-project configuration is needed. The registration is equivalent to running:

claude mcp add --scope user tigermemory npx tigermemory remote-client

This makes Tiger Memory available in all Claude Code sessions automatically.

🔒 Privacy & Security

• GitHub OAuth Authentication: Secure authentication via GitHub with proper scope handling
• You control sharing: Decisions marked public: false stay private to your project
• Secure API keys: Authentication tokens stored locally and transmitted securely
• Secure connections: All connections use HTTPS/WSS with proper encryption
• Data isolation: User-based access control with team collaboration support
• Audit trail: Complete timeline of all decisions with user attribution

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

Development Setup

# Clone the repository
git clone https://github.com/tigermemory/tigermemory.git
cd tigermemory

# Install dependencies
npm install

# Set up environment
cp .env.example .env
# Edit .env with your credentials

# Run in development mode
npm run dev

# Run tests
npm test

# Build for production
npm run build

🙋‍♀️ Support

• Documentation: Tiger Memory Docs
• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Email: [email protected]

🗺️ Roadmap

• [x] Multi-User Support - GitHub OAuth authentication with user attribution
• [x] Remote MCP Server - Cloud-native SSE transport for seamless integration
• [ ] Team Collaboration - Enhanced team features and project sharing
• [ ] Pattern Marketplace - Community-driven architectural pattern sharing
• [ ] IDE Extensions - Direct integration with VS Code, IntelliJ, etc.
• [ ] Advanced Analytics - Decision success tracking and recommendations
• [ ] Multi-LLM Support - Support for other AI coding assistants
• [ ] Self-Hosted Option - Run your own Tiger Cloud instance

Tiger Memory - Making Claude Code remember, so you don't have to repeat yourself.

Built with ❤️ for the developer community