ContextKeep

Provides infinite, local memory for AI tools through MCP with local or SSE-based access.

python

0

GitHub Stars

python

Language

5 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

ContextKeep MCP Server provides a private, persistent memory store for AI agents. It lets you save unlimited contexts, memories, and preferences locally or remotely, enabling faster, more relevant retrieval and reduced token usage across your AI tools.

How to use

You connect to ContextKeep using an MCP client in one of two ways: a local stdio configuration for direct integration with your AI tools, or a remote HTTP (SSE) transport for web apps and other clients. Once connected, you can store memories, search semantically, and retrieve only the most relevant context to use with your AI tool.

Typical usage patterns include:

Persisting project notes, preferences, and snippets for long-term recall across sessions.
Performing semantic searches to quickly locate relevant memories before you run prompts.
Managing memories through a simple CRUD interface in the web dashboard for manual edits.

Notes on usage with MCP clients

If you are integrating as a local process, you will run a Python-based server in a virtual environment and point your MCP client at the server process. If you prefer HTTP transport, you can connect via the built-in SSE endpoint exposed by the server.