LycheeMem

LycheeMem

A lightweight long-term memory framework for LLM agents.

MCPDevOpen source
Type
MCP
Transport
http
Open source
Yes
GitHub Stars
★ 1.0k
Source
mcp-github

Overview

LycheeMemory is a compact memory framework designed for LLM agents. It starts with efficient conversation memory and progressively expands into more powerful action-aware and usage-aware memory systems through structured organization, lightweight integration, and adaptive retrieval. This capability can be integrated via multiple methods, including OpenClaw plugin, Claude Code plugin, Hermes plugin, PyPI package, and HTTP MCP server. Ideal for developers and researchers seeking enhanced memory functionality.

Capabilities

  • Efficient conversation memory
  • Structured organization
  • Lightweight integration
  • Adaptive retrieval
  • Action-aware memory
  • Usage-aware memory

Use cases

Enhance memory capabilities of LLM agentsReduce token consumption and costsImprove accuracy of semantic memory searchSupport multimodal memory

Setup

Requires: Python 3.9+LanceDB
pip install lycheemem; use lycheemem-cli to start the service

This information was compiled by AI from public sources and may contain inaccuracies — please refer to the source.

FAQ

How to install LycheeMemory?

Install using pip install lycheemem.

Which runtime environments are supported?

Supports OpenClaw, Claude Code, Hermes, and others.

Is an additional database required?

Yes, LanceDB is required as the vector database.

Related skills