EverOS

EverOS

A self-evolving memory layer for multi-agent and platform systems.

MCPDevOpen source
Type
MCP
Open source
Yes
GitHub Stars
★ 8.5k
Source
mcp-github

Overview

EverOS is a Python library and local-first memory runtime designed for agents and developers. It provides a portable memory layer from the outset for coding assistants, applications, devices, and workflows. It stores conversations, files, and agent trajectories in human-readable Markdown format and synchronizes local SQLite and LanceDB indexes for fast retrieval and self-evolving reuse. With EverOS, AI gains a consistent memory experience across multiple agents and platforms, enhancing performance across diverse scenarios.

Capabilities

  • Memory storage in Markdown format
  • Direct editing of Markdown files
  • Local three-part stack (Markdown + SQLite + LanceDB)
  • User and agent-independent memory trajectories
  • Orthogonal retrieval capabilities

Use cases

Memory sharing across multiple agents and platformsImproved AI performance across different contextsSimplified memory management and retrievalSupport for multimodal chat and embedded search

Setup

Requires: Python 3.12+API Key
pip install everos

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

FAQ

How do I get started with EverOS?

Run `everos demo` after installation to see a demonstration, then configure your API keys and start the server.

What API keys are required?

You need API keys for OpenRouter and DeepInfra.

Related skills