What it does
The server exposes nine tools to an LLM: add_memory, search_memories, get_memories, get_memory, update_memory, delete_memory, delete_all_memories, delete_entities and list_entities. They cover saving text or conversation history, semantic search, paginated listing, retrieving and updating single memories, and deleting memories at several scopes. Responses are JSON strings returned directly from the Mem0 API.
It suits MCP-compatible clients such as Claude Desktop, Cursor or custom agents that need persistent memory of users, for example storing health details, research parameters or project status and retrieving them later. It can run locally as a Python package started with uvx or pip, as a Docker container exposing an HTTP endpoint, or connected to a Smithery-hosted remote server. The repository also includes a bundled Pydantic AI test agent and source setup instructions.
Running it requires a Mem0 platform API key set as MEM0_API_KEY. Optional variables set a default user ID, enable graph memories by default and choose the model used by the bundled test agent. The project is released under the Apache License 2.0.
Limitations and requirements
The publisher states the project has been archived, is no longer actively maintained and is now a public archive, and points to an official cloud-hosted MCP server instead. A Mem0 platform API key (MEM0_API_KEY) is required.
Safety
Our team has not reviewed this resource. Only automated rules checked it; resources with high-risk signals are not published. These are signals, not a guarantee that installing it is safe. Try it in a sandbox project first.
- Runs code or changes an assistant's behaviour
Sources
- Cline MCP Marketplace first seen 13 Nov 2026 · last seen 18 Nov 2026 · its page on the source
661
As shown on cline.bot when we last read it