Mcp Memory Bank
@bsmi021
A powerful, production-ready context management system for Large Language Models (LLMs). Built with ChromaDB and modern embedding technologies, it provides persistent, project-specific memory capabilities that enhance your AI's understanding and response quality.
Overview
What is Mcp Memory Bank?
—
How to use Mcp Memory Bank?
Deploy using Docker Compose. Run docker-compose up --build -d to build and start the application and its required ChromaDB vector database.
Key features of Mcp Memory Bank
- Docker-ready deployment with Docker Compose
- Uses ChromaDB as vector database
- Configurable embedding model via environment variable
- Persistent data volumes for ChromaDB
- HTTP API on port 3000
- Non-root user inside container
Use cases of Mcp Memory Bank
—
FAQ from Mcp Memory Bank
What environment variables are required?
Default environment variables include CHROMADB_URL, TRANSPORT, HTTP_PORT, MCP_MEMBANK_EMBEDDING_MODEL, NODE_ENV, and NODE_OPTIONS. They can be overridden in your environment or in docker-compose.yml.
What is the default embedding model?
The default embedding model is Xenova/all-MiniLM-L6-v2, configurable via MCP_MEMBANK_EMBEDDING_MODEL.
Is ChromaDB data persisted?
Yes, ChromaDB data is persisted in the named Docker volume chromadb-data.
What ports does the application use?
Port 3000 for the main application HTTP API, and port 8000 for ChromaDB.
Does the application run as root?
No, the application runs inside the container as a non-root user.
