Mem Agent
mem-agent mcp server
What is Mem Agent?
Mem Agent is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to mem-agent mcp server
mem-agent mcp server
This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- mem-agent mcp server
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx mem-agentConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Mem Agent
Mem Agent MCP is a local AI memory system that gives language models persistent, structured long-term memory through a markdown-based knowledge graph stored on disk. It runs a fine-tuned local model (mem-agent-mlx-4bit or mem-agent-mlx-8bit) via LM Studio and exposes MCP tools for writing and retrieving memories, filtering sensitive information, and importing conversation history from ChatGPT, Notion, Nuclino, GitHub, and Google Docs. Developers and researchers who need their AI assistant to remember personal context, project knowledge, and entity relationships across sessions use Mem Agent to build a persistent, privacy-respecting memory layer without sending data to cloud services.
Prerequisites
- macOS with Apple Silicon (MLX models require Metal acceleration)
- LM Studio installed (the setup Makefile installs it automatically on macOS)
- uv Python package manager installed
- A memory storage directory (any local folder) designated during setup
- An MCP-compatible AI client such as Claude Desktop or Claude Code
Install LM Studio and verify uv
Use the Makefile targets to install LM Studio on macOS and confirm uv is available before proceeding.
make check-uv
make installSet up the memory directory and select a model
Run the setup target to choose your memory storage directory. You will be prompted to select model precision (4-bit or 8-bit) during the next step.
make setupGenerate the MCP configuration file
The generate-mcp-json target creates the MCP server configuration file with the correct paths and model settings.
make generate-mcp-jsonImport existing memory sources (optional)
Use the memory wizard to import conversation history from ChatGPT exports, Notion/Nuclino ZIP files, GitHub (PAT required), or Google Docs (OAuth token required).
make memory-wizard
# For manual connector CLI:
make connect-memoryAdd custom privacy filters (optional)
Configure filters to prevent Mem Agent from revealing or storing sensitive information categories in responses.
make add-filters
# To clear all filters:
make reset-filtersStart the MCP server
Launch Mem Agent over STDIO (for Claude Desktop) or HTTP (for remote access). STDIO is recommended for local use with Claude Desktop.
make serve-mcp
# Or for HTTP transport:
make serve-mcp-httpMem Agent Examples
Client configuration
Use the generated MCP JSON config (from make generate-mcp-json) or manually add the server entry to claude_desktop_config.json.
{
"mcpServers": {
"mem-agent": {
"command": "uv",
"args": ["run", "serve-mcp"],
"env": {
"VLLM_HOST": "localhost",
"VLLM_PORT": "4000"
}
}
}
}Prompts to try
Prompts that demonstrate Mem Agent's persistent memory and privacy filtering capabilities.
- "What do you remember about my current project roadmap?"
- "What were my thoughts on AI agent frameworks from our past conversations?"
- "Tell me about the relationship between Alice and the backend team <filter> Do not reveal personal contact information"
- "Summarise my recent technical discussions about database architecture"
- "What is the background of the entity called ProjectPhoenix?"Troubleshooting Mem Agent
LM Studio model fails to load or returns an error about Metal acceleration
Mem Agent's MLX models require Apple Silicon (M1/M2/M3). Ensure you are on macOS with an Apple Silicon Mac and that LM Studio has been granted the necessary permissions to use Metal in System Settings.
Memory retrieval returns irrelevant or empty results
Check that your memory directory contains the expected markdown files in the correct structure (user.md at root, entities/ subdirectory with entity files). Wikilinks between files (e.g. [[entities/Alice.md]]) are required for relationship traversal.
GitHub or Google Docs connector authentication fails during import
For GitHub, generate a Personal Access Token with public_repo or repo scope. For Google Docs, obtain an OAuth 2.0 access token with the drive.readonly scope. Pass these during 'make connect-memory' when prompted.
Frequently Asked Questions about Mem Agent
What is Mem Agent?
Mem Agent is a Model Context Protocol (MCP) server that mem-agent mcp server It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Mem Agent?
Follow the installation instructions on the Mem Agent GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Mem Agent?
Mem Agent works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Mem Agent free to use?
Yes, Mem Agent is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.
Mem Agent Alternatives — Similar Knowledge & Memory Servers
Looking for alternatives to Mem Agent? 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 Mem Agent 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 Mem Agent?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.