OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Memory & RAG

Reme

Reme is an open-source Python memory-management toolkit that stores durable agent knowledge as readable, editable Markdown files in a local workspace. It provides file operations and retrieval using BM25 search and wikilink expansion, with optional embeddings and workflows for evolving memory over time.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Reme — OpenViST project profile in Agent Memory & RAG.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Reme when an agent needs persistent, inspectable memory across conversations, or when multiple agents should work from the same Markdown knowledge base.

Who it is for

  • Developers adding durable memory to an AI agent.
  • People building local knowledge bases from agent conversations and other resources.
  • Teams whose agents need to share and maintain a persistent memory workspace.

Documented uses

  • Convert conversations and resources into daily notes and longer-term Markdown knowledge.
  • Search memory for relevant passages and follow connections between related files.
  • Inspect, edit, move, sync, or back up memory files using familiar file tools.
  • Connect an agent to a shared workspace through CLI, HTTP, MCP, or the Python API.
Before you choose

Requirements & limits

Reme is open-source under the Apache-2.0 license. The core package requires Python 3.11 or newer.

  • Choose an integration path based on the host agent: Reme documents CLI, HTTP, MCP, and Python API access, plus integrations for DeepSeek Harness, OpenClaw, and QwenPaw.
  • Basic file operations, BM25 search, and wikilink traversal do not require LLM credentials; automatic memory evolution workflows do.
  • Embedding-based retrieval is supported but disabled by default and requires enabling the related components in configuration.
  • The optional Studio interface is included with the core installation; a base installation can add it with the web extra.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install Reme and start its local service

Install the core package and launch Reme’s local service.

Before you start

  • Python 3.11 or newer

Steps

  1. Install the core package with the command below.

    Documented command
    pip install "reme-ai[core]"

  2. Run reme start.

  3. Use the default service address, 127.0.0.1:2333, or set a different port with reme start service.port=8181.

Expected outcome

The Reme service starts locally; its default workspace is created under `.reme/` in the current directory.

Guide limitations

  • This setup guide does not configure an agent integration.
  • LLM credentials are required for LLM-powered workflows such as auto_memory, auto_resource, auto_dream, and proactive refresh.
  • Embedding retrieval is disabled by default.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
1 October 2026
Original website title
agentscope-ai/ReMe · Official homepage title · retained in its source language
How we review sourcesSuggest a correction