Show HN: A Real and Proactive MCP Memory Tool

1 day ago 5

Smart memory for AI assistants - A Model Context Protocol (MCP) server that remembers your conversations, learns from patterns, and provides intelligent context suggestions.

Perfect for Claude Desktop, VS Code, Continue, Cursor, and any MCP-compatible AI client.

🚀 Quick Start (5 minutes)

Option 1: Single Docker Commands (Quick Test)

# 1. Start Chroma vector database docker run -d --name mcp-chroma \ -p 8000:8000 \ -v chroma_data:/chroma/chroma \ chromadb/chroma:latest # 2. Start MCP Memory Server docker run -d --name mcp-memory \ -p 9080:9080 \ -e OPENAI_API_KEY="your-api-key-here" \ -e MCP_MEMORY_CHROMA_ENDPOINT="http://host.docker.internal:8000" \ -v mcp_data:/app/data \ --link mcp-chroma:chroma \ ghcr.io/fredcamaral/mcp-memory:latest # 3. Optional: Auto-updater (watches for new releases) docker run -d --name mcp-auto-updater \ -v /var/run/docker.sock:/var/run/docker.sock \ -e WATCHTOWER_CLEANUP=true \ -e WATCHTOWER_POLL_INTERVAL=3600 \ -e WATCHTOWER_SCOPE=mcp-memory \ containrrr/watchtower:latest

Option 2: Docker Compose (Recommended - Full Setup)

Clone and start everything:

git clone https://github.com/fredcamaral/mcp-memory.git cd mcp-memory cp .env.example .env # Edit .env and add your OPENAI_API_KEY # Local development (builds from source) docker-compose up -d # Production with auto-updates (uses registry + Watchtower) docker-compose -f docker-compose.yml -f docker-compose.prod.yml --profile auto-update up -d
Configure your AI client (e.g., Claude Desktop, Claude Code, Windsurf, Cursor, etc):

!!! The MCP works through a simple stdio <> HTTP proxy written in javascript, as per shown below. !!! SSE protocol is also available @ :9080/sse

{ "mcpServers": { "memory": { "type": "stdio", "command": "docker", "args": ["exec", "-i", "mcp-memory-server", "node", "/app/mcp-proxy.js"] } } }
Test it! 🎉
- Open your AI client (Claude Desktop, etc.)
- Ask it to store a memory: "Please remember that I prefer TypeScript over JavaScript"
- Later ask: "What do you remember about my coding preferences?"

📋 Full docker-compose.yml Reference

services: # Chroma Vector Database chroma: image: chromadb/chroma:latest container_name: mcp-chroma restart: unless-stopped ports: - "8000:8000" volumes: - chroma_data:/chroma/chroma networks: - mcp_network # Main MCP Memory Server mcp-memory-server: build: context: . dockerfile: Dockerfile container_name: mcp-memory-server restart: unless-stopped depends_on: - chroma ports: - "9080:9080" - "9081:8081" - "9082:8082" environment: - OPENAI_API_KEY=${OPENAI_API_KEY} - MCP_MEMORY_CHROMA_ENDPOINT=http://chroma:8000 - MCP_MEMORY_DB_TYPE=sqlite - MCP_MEMORY_DB_PATH=/app/data/memory.db volumes: - mcp_data:/app/data - mcp_logs:/app/logs - mcp_backups:/app/backups networks: - mcp_network # Auto-updater sidecar (optional - use --profile auto-update) watchtower: image: containrrr/watchtower:latest container_name: mcp-auto-updater restart: unless-stopped environment: - WATCHTOWER_CLEANUP=true - WATCHTOWER_POLL_INTERVAL=3600 # Check hourly - WATCHTOWER_SCOPE=mcp-memory-server volumes: - /var/run/docker.sock:/var/run/docker.sock networks: - mcp_network profiles: - auto-update networks: mcp_network: driver: bridge volumes: chroma_data: name: mcp_memory_chroma_vector_db_NEVER_DELETE mcp_data: name: mcp_memory_app_data_NEVER_DELETE mcp_logs: name: mcp_memory_logs_NEVER_DELETE mcp_backups: name: mcp_memory_backups_NEVER_DELETE

MCP Memory transforms your AI assistant into a smart companion that:

📚 Remembers Everything: Stores all your conversations and contexts across sessions
🔍 Smart Search: Finds relevant past conversations using AI-powered similarity search
🧠 Pattern Learning: Recognizes your preferences, coding patterns, and decision-making
💡 Proactive Suggestions: Automatically suggests relevant context from your history
🔄 Cross-Project Intelligence: Learns patterns across all your repositories and projects

Claude Desktop Configuration

Configuration:

{ "mcpServers": { "memory": { "type": "stdio", "command": "docker", "args": ["exec", "-i", "mcp-memory-server", "node", "/app/mcp-proxy.js"] } } }

Add to your Continue configuration:

{ "models": [...], "mcpServers": { "memory": { "type": "stdio", "command": "docker", "args": ["exec", "-i", "mcp-memory-server", "node", "/app/mcp-proxy.js"] } } }

Environment Variables (.env file)

# Required OPENAI_API_KEY=your-openai-api-key-here # Optional (defaults work for Docker setup) CHROMA_URL=http://chroma:8000 MCP_MEMORY_DATA_DIR=./data MCP_MEMORY_LOG_LEVEL=info

Memory Tools Available to Your AI

Once configured, your AI assistant automatically gets these powerful memory abilities:

Store important moments: memory_store_chunk - Save conversations, decisions, solutions
Smart search: memory_search - Find similar past conversations and contexts
Get context: memory_get_context - Retrieve project overview and recent activity
Find patterns: memory_get_patterns - Identify recurring themes and solutions
Health monitoring: memory_health_dashboard - Track memory system effectiveness
Intelligent decay: memory_decay_management - Automatically summarize and archive old memories

🧠 Conversation Flow Detection: Recognizes when you're debugging, implementing, or planning
🔗 Relationship Mapping: Automatically links related memories and contexts
📊 Pattern Recognition: Learns your coding patterns, preferences, and decision-making
💡 Smart Suggestions: Proactively suggests relevant memories based on current context
🗂️ Multi-Repository Support: Works across all your projects with intelligent cross-referencing

🔴 "Connection refused" or "Server not responding"

# Check if containers are running docker-compose ps # Check logs docker-compose logs mcp-memory-server # Restart services docker-compose restart

🔴 "OpenAI API errors"

Check your API key in .env file
Verify you have credits in your OpenAI account
Check network connectivity

🔴 "Memory not persisting"

# Check database connection docker-compose logs chroma # Verify data directory permissions ls -la ./data/

Checking if Everything Works

Test the server directly:

curl http://localhost:8081/health
Browse the web interface:
- Open http://localhost:8082 in your browser
- You should see the memory management dashboard
Test with your AI client:
- Ask it to remember something: "Please store that I work on the mcp-memory project"
- Ask it to recall: "What do you remember about my current projects?"

🎛️ Advanced Configuration

For production use, see the detailed configurations:

# configs/custom/config.yaml storage: chroma: url: "http://your-chroma-instance:8000" embeddings: openai: api_key: "${OPENAI_API_KEY}" model: "text-embedding-ada-002" security: encryption: enabled: true access_control: enabled: true