mem0-mcp-server

MCP server exposing Mem0 v2 API for AI agents to store, retrieve, and search long-term memories using semantic search through the standardized MCP protocol.

Mem0-MCP Server is a self-hosted MCP (Model Context Protocol) server that bridges AI agents with persistent memory storage. It enables intelligent context retention across conversations and sessions using Mem0's AsyncMemory API.

- MCP Protocol Integration - Exposes Mem0 functionality via MCP tools
- Semantic Memory Search - Similarity-based memory retrieval with vector search
- Multi-Tenant Isolation - User/Agent/Session scoped memory isolation
- Flexible Transport - stdio for local agents, SSE for remote connections
- Configuration Management - Pydantic-based validation with environment variable support

# Clone and install git clone https://github.com/your-org/mem0-mcp-server.git cd mem0-mcp-server uv sync # Set environment variables export OPENAI_API_KEY="your-api-key"

Create~/.config/mem0-mcp-server/settings.json:

{ "vector_store": { "provider": "redis", "config": { "redis_url": "redis://localhost:6379" } }, "llm": { "provider": "openai", "config": { "model": "gpt-4o" } }, "embedder": { "provider": "openai", "config": { "model": "text-embedding-3-small" } } }

# SSE Transport (remote connections) uv run python -m mcp_server.main # stdio Transport (local AI agents) export MCP_TRANSPORT=stdio uv run python -m mcp_server.main

"mcp": { "mem0": { "type": "remote", "enabled": true, "url": "http://localhost:8050/sse" } }

# Add memory result = await client.call_tool("add_memory", { "messages": [{"role": "user", "content": "I prefer dark mode"}], "user_id": "alice" }) # Search memories result = await client.call_tool("search_memories", { "query": "theme preferences", "filters": {"user_id": "alice"}, "limit": 5 })

AI Agent → FastMCP Server → MemoryManager → Mem0 AsyncMemory → Redis │ │ ├── SafeLogger (stdout/stderr) │ ├── Transport (stdio/SSE) │ └── Config (Pydantic validation) │

- COMP-1: ConfigLoader - Configuration loading and validation
- COMP-2: FastMCP Server - MCP protocol server
- COMP-3: MemoryManager - Memory operations with multi-tenant isolation
- COMP-4: MCP Tools - Tool definitions
- COMP-5: SafeLogger - Output stream separation

Configuration values are resolved in order:
- Tool parameters (direct)
- Environment variables (with MCP_ prefix)
- Config file values
- Hardcoded defaults

# Using docker-compose docker-compose up -d # Using Makefile make docker-up # Start services with docker compose make docker-down # Stop services make docker-logs # Show logs

SeeDeployment → Dockerfor detailed configuration.

# Using Helm chart helm install mem0-mcp ./charts/mem0-mcp-server

SeeMakefilefor all available commands including Docker management (docker-up,docker-down,docker-logs, etc.).

mem0-mcp/ ├── src/mcp_server/ │ ├── __init__.py # FastMCP singleton │ ├── lifespan.py # Resource lifecycle │ ├── transport.py # Transport selection │ ├── memory/ │ │ ├── manager.py # MemoryManager │ │ └── lifespan.py # AsyncMemory lifecycle │ ├── config/ │ │ ├── settings.py # Pydantic models │ │ └── loader.py # Config file loading │ ├── tools/ │ │ ├── add_memory.py │ │ ├── search_memories.py │ │ └── ... │ └── utils/ │ └── safe_logger.py # Output separation ├── doc/ │ ├── api/ # API reference │ ├── patterns/ # Pattern guides │ ├── examples/ # Usage examples │ └── architecture/ # Architecture docs ├── tests/ ├── Makefile ├── Dockerfile └── docker-compose.yml

Auditable, self-improving knowledge & memory for AI agents over MCP — citation-enforced answers and a replayable why-trace, self-hosted on Postgres.