Uses SQLite for persistent memory storage, enabling the creation, retrieval, updating, and deletion of memories with features like titled entries and memory management through various tools.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Memory MCP remember meeting notes from yesterday's project kickoff"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A Model Context Protocol server for storing and retrieving memories using low-level Server implementation and SQLite storage.
This project uses uv for dependency management instead of pip. uv is a fast, reliable Python package installer and resolver.
Install using uv:
uv pip install memory-mcpOr install directly from source:
uv pip install .For development:
uv pip install -e ".[dev]"If you don't have uv installed, you can install it following the official instructions.
Related MCP server: SQLite MCP Server
memory-mcpThis will start the MCP server that allows you to store and retrieve memories.
The Memory MCP provides the following tools:
remember: Store a new memory with a title and content
get_memory: Retrieve a specific memory by ID or title
list_memories: List all stored memories
update_memory: Update an existing memory
delete_memory: Delete a memory
MCP provides a handy command-line tool called mcp inspect that allows you to debug and interact with your MCP server directly.
First, make sure the MCP CLI tools are installed:
uv pip install mcp[cli]Start the Memory MCP server in one terminal:
memory-mcpIn another terminal, connect to the running server using mcp inspect:
mcp inspectOnce connected, you can:
> toolsThis will display all the tools provided by the Memory MCP server.
To call a tool, use the call command followed by the tool name and any required arguments:
> call remember title="Meeting Notes" content="Discussed project timeline and milestones."> call list_memories> call get_memory memory_id=1> call update_memory memory_id=1 title="Updated Title" content="Updated content."> call delete_memory memory_id=1You can enable debug mode to see detailed request and response information:
> debug onThis helps you understand exactly what data is being sent to and received from the server.
To view the schema for a specific tool:
> tool rememberThis shows the input schema, required parameters, and description for the tool.
If you encounter issues:
Check the server logs in the terminal where your server is running for any error messages.
In the MCP inspect terminal, enable debug mode with debug on to see raw requests and responses.
Ensure the tool parameters match the expected schema (check with the tool command).
If the server crashes, check for any uncaught exceptions in the server terminal.
To contribute to the project, install the development dependencies:
uv pip install -e ".[dev]"This project uses uv.lock file to lock dependencies. To update dependencies:
uv pip compile pyproject.toml -o uv.lockpython -m pytestblack memory_mcp testsruff check memory_mcp testsmypy memory_mcpDelete a memory.
Inputs: memory_id (required).
Retrieve a specific memory by ID or title.
Inputs: memory_id, title.
List all stored memories.
Inputs: no named parameters.
Store a new memory.
Inputs: content (required), title (required).
Update an existing memory.
Inputs: memory_id (required), content, title.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 5 tools
Each tool has a clearly distinct purpose targeting specific CRUD operations on memories: list (retrieve all), get (retrieve specific), remember (create), update (modify), and delete (remove). There is no overlap in functionality, making tool selection unambiguous for an agent.
All tool names follow a consistent verb_noun pattern with snake_case, using clear action verbs (list, get, remember, update, delete) paired with the noun 'memory'. There are no deviations or mixed conventions, ensuring predictable naming throughout.
With 5 tools, this server is well-scoped for memory management, providing complete CRUD coverage without unnecessary bloat. Each tool earns its place by covering essential operations, making the count appropriate for the domain.
The tool set offers complete CRUD/lifecycle coverage for memory management: create (remember), read (list_memories, get_memory), update (update_memory), and delete (delete_memory). There are no obvious gaps, allowing agents to handle all core workflows without dead ends.
Correction-first memory for AI agents. Run it locally or connect any MCP agent to the hosted server.
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
Hosted memory for AI agents over MCP. One memory shared by all your agents.
An MCP memory server. One memory your agents share — across models, devices and apps.