Dragon Brain
Dragon Brain — persistent long-term memory for AI agents via MCP (Model Context Protocol). Knowledge graph (FalkorDB) + vector search (Qdrant) + CUDA GPU embeddings. Works with Claude, Gemini CLI, Cursor, Windsurf, VS Code Copilot. 30 tools, 1121 tes
What is Dragon Brain?
Dragon Brain is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to dragon brain — persistent long-term memory for ai agents via mcp (model context protocol). knowledge graph (falkordb) + vector search (qdrant) + cuda gpu embeddings. works with claude, gemini cli, cur...
Dragon Brain — persistent long-term memory for AI agents via MCP (Model Context Protocol). Knowledge graph (FalkorDB) + vector search (Qdrant) + CUDA GPU embeddings. Works with Claude, Gemini CLI, Cursor, Windsurf, VS Code Copilot. 30 tools, 1121 tes
This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Dragon Brain — persistent long-term memory for AI agents via
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx dragon-brainConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Dragon Brain
Dragon Brain is a production-grade persistent memory MCP server for AI agents, combining a FalkorDB knowledge graph with Qdrant vector search and BGE-M3 embeddings to achieve 100% recall@5 on the LongMemEval benchmark. It exposes 34 MCP tools covering entity storage, semantic search, relationship traversal, time-travel queries, and autonomous memory clustering — all without requiring an LLM for retrieval. Works with Claude Code, Claude Desktop, Cursor, Windsurf, Gemini CLI, and any MCP-compatible client.
Prerequisites
- Docker and Docker Compose installed
- Python 3.12 or higher (if using the pip package instead of Docker-only mode)
- An MCP-compatible client such as Claude Code, Claude Desktop, or Cursor
- NVIDIA GPU with CUDA drivers (optional, for GPU-accelerated embeddings)
Start the backend services with Docker Compose
Dragon Brain requires four Docker containers: FalkorDB (knowledge graph on port 6379), Qdrant (vector search on port 6333), an Embedding API serving BGE-M3 on port 8001, and a Streamlit dashboard on port 8501. Start them all with one command.
# CPU mode (default)
docker compose up -d
# GPU mode (NVIDIA CUDA)
docker compose --profile gpu up -dVerify all containers are healthy
Check that all four containers are running before connecting your MCP client.
docker ps --filter "name=claude-memory"Install the MCP server package
Install the dragon-brain Python package which provides the MCP server binary.
pip install dragon-brainConnect Claude Code (recommended method)
Register Dragon Brain with Claude Code using the mcp add command.
claude mcp add dragon-brain -- python -m claude_memory.serverAdd to Claude Desktop or other MCP clients
For Claude Desktop or other clients, add the server configuration with the service connection environment variables.
{
"mcpServers": {
"dragon-brain": {
"command": "python",
"args": ["-m", "claude_memory.server"],
"env": {
"FALKORDB_HOST": "localhost",
"FALKORDB_PORT": "6379",
"QDRANT_HOST": "localhost",
"QDRANT_PORT": "6333",
"EMBEDDING_API_URL": "http://localhost:8001"
}
}
}
}Start storing and retrieving memories
Tell your AI assistant to remember things. Dragon Brain creates typed entities, observations, and relationships in the knowledge graph and indexes them for semantic search across all future sessions.
Dragon Brain Examples
Client configuration
Claude Desktop configuration for Dragon Brain with local Docker services.
{
"mcpServers": {
"dragon-brain": {
"command": "python",
"args": ["-m", "claude_memory.server"],
"env": {
"FALKORDB_HOST": "localhost",
"FALKORDB_PORT": "6379",
"QDRANT_HOST": "localhost",
"QDRANT_PORT": "6333",
"EMBEDDING_API_URL": "http://localhost:8001"
}
}
}
}Prompts to try
Example prompts to store and retrieve memories using Dragon Brain's 34 tools.
- "Remember that I'm building Atlas in Rust and I prefer functional patterns"
- "What do you know about my projects?"
- "What relationships connect Project Atlas to distributed systems in your memory?"
- "What did you know about my preferences last Tuesday?"
- "Find all entities related to machine learning in my memory graph"
- "Search my memories for anything about database architecture"Troubleshooting Dragon Brain
Docker containers fail to start or show as unhealthy
Run `docker compose logs` to see per-container errors. Common causes: port 6379 or 6333 already in use by another service, or insufficient Docker memory allocation (increase to at least 4GB in Docker Desktop settings).
Embedding API container is slow or times out
The BGE-M3 embedding model (~1GB) downloads on first start. Wait for the container to finish pulling and initializing before sending requests. Check progress with `docker logs claude-memory-embedding-1`.
claude mcp add succeeds but the server doesn't appear in Claude Code
Ensure the Docker services are running first (`docker compose up -d`) before starting Claude Code. Dragon Brain's server fails loudly if FalkorDB or Qdrant are unreachable — check that all four containers are healthy.
Frequently Asked Questions about Dragon Brain
What is Dragon Brain?
Dragon Brain is a Model Context Protocol (MCP) server that dragon brain — persistent long-term memory for ai agents via mcp (model context protocol). knowledge graph (falkordb) + vector search (qdrant) + cuda gpu embeddings. works with claude, gemini cli, cursor, windsurf, vs code copilot. 30 tools, 1121 tes It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Dragon Brain?
Follow the installation instructions on the Dragon Brain GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Dragon Brain?
Dragon Brain works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Dragon Brain free to use?
Yes, Dragon Brain is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Dragon Brain Alternatives — Similar Knowledge & Memory Servers
Looking for alternatives to Dragon Brain? Here are other popular knowledge & memory servers you can use with Claude, Cursor, and VS Code.
MemPalace
★ 52.6kA local AI memory system that stores all conversations verbatim and organizes them into navigable structures. It provides 19 MCP tools for AI assistants to search and retrieve past decisions, debugging sessions, and architecture debates automatically
Kratos
★ 25.7k🏛️ Memory System for AI Coding Tools - Never explain your codebase again. MCP server with perfect project isolation, 95.8% context accuracy, and the Four Pillars Framework.
Context Mode
★ 15.4kAn MCP server that preserves LLM context by intercepting large data outputs and returning only concise summaries or relevant sections. It enables efficient sandboxed code execution, file processing, and documentation indexing across multiple programm
Memu
★ 13.7kMemory for 24/7 proactive agents like OpenClaw.
MemOS
★ 9.3kMemOS (Memory Operating System) is a memory management operating system designed for AI applications. Its goal is: to enable your AI system to have long-term memory like a human, not only remembering what users have said but also actively invoking, u
Everos
★ 5.4kBuild, evaluate, and integrate long-term memory for self-evolving agents.
Browse More Knowledge & Memory MCP Servers
Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Dragon Brain in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Dragon Brain?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.