MetaClaw

<div align="center">

<img src="assets/new_logo2.png" alt="MetaClaw" width="600">

<br/>

# Just talk to your agent — it learns and *EVOLVES*.

<p>Inspired by how brains learn. Meta-learn and evolve your 🦞 from every conversation in the wild. No GPU required.
  
<br/>


<img src="assets/metaclaw_mainfig_v2.png" alt="MetaClaw Architecture" width="800">

<br/>


<p>
  <a href="https://arxiv.org/abs/2603.17187"><img src="https://img.shields.io/badge/📄_Technical_Report-purple?style=flat-square" alt="Tech Report" /></a>
  <a href="https://github.com/aiming-lab/MetaClaw"><img src="https://img.shields.io/badge/github-MetaClaw-181717?style=flat&labelColor=555&logo=github&logoColor=white" alt="GitHub"></a>
  <a href="LICENSE"><img src="https://img.shields.io/badge/License-MIT-green?style=flat&labelColor=555" alt="License MIT"></a>
  <img src="https://img.shields.io/badge/⚡_Fully_Async-yellow?style=flat&labelColor=555" alt="Fully Async" />
  <img src="https://img.shields.io/badge/☁️_No_GPU_Cluster-blue?style=flat&labelColor=555" alt="No GPU Cluster" />
  <img src="https://img.shields.io/badge/🛠️_Skill_Evolution-orange?style=flat&labelColor=555" alt="Skill Evolution" />
  <img src="https://img.shields.io/badge/🚀_One--Click_Deploy-green?style=flat&labelColor=555" alt="One-Click Deploy" />
</p>

[🇨🇳 中文](./assets/README_ZH.md) • [🇯🇵 日本語](./assets/README_JA.md) • [🇰🇷 한국어](./assets/README_KO.md) • [🇫🇷 Français](./assets/README_FR.md) • [🇩🇪 Deutsch](./assets/README_DE.md) • [🇪🇸 Español](./assets/README_ES.md) • [🇧🇷 Português](./assets/README_PT.md) • [🇷🇺 Русский](./assets/README_RU.md) • [🇮🇹 Italiano](./assets/README_IT.md) • [🇻🇳 Tiếng Việt](./assets/README_VI.md) • [🇦🇪 العربية](./assets/README_AR.md) • [🇮🇳 हिन्दी](./assets/README_HI.md)

<br/>

[Overview](#-overview) • [Quick Start](#-quick-start) • [Multi-Claw Support](#-multi-claw-support) • [Configuration](#️-configuration) • [Skills Mode](#-skills-mode) • [RL Mode](#-rl-mode) • [MadMax Mode](#-madmax-mode-default) • [Memory](#-memory) • [Citation](#-citation)

</div>

---

<div align="center">

### Two commands. That's it.
</div>

```bash
metaclaw setup              # one-time config wizard
metaclaw start              # default: madmax mode — skills + scheduled RL training
metaclaw start --daemon     # run in background, logs -> ~/.metaclaw/metaclaw.log
metaclaw start --daemon --log-file /tmp/metaclaw.log  # custom daemon log path
metaclaw start --mode rl    # RL without scheduler (trains immediately on full batch)
metaclaw start --mode skills_only  # skills only, no RL (no Tinker needed)
```

<div align="center">
<img src="assets/metaclaw.gif" alt="MetaClaw demo" width="700">
</div>

---

## 🔥 News

- **[03/25/2026]** **v0.4.0** — Contexture layer: MetaClaw now persists cross-session memory for users and projects. Relevant facts, preferences, and project history are automatically retrieved and injected into prompts. Includes adaptive memory policy, background consolidation, and an optional memory sidecar service.
- **[03/24/2026]** **v0.3.3** — One-click OpenClaw plugin: MetaClaw now ships as a native OpenClaw extension — drop the folder into OpenClaw's extensions, run one command, and everything is set up automatically.
- **[03/18/2026]** Our technical report "[MetaClaw: Just Talk -- An Agent That Meta-Learns and Evolves in the Wild](https://arxiv.org/pdf/2603.17187)" is out! **🏆 Ranked No. 1** on [HuggingFace Daily Papers](https://huggingface.co/papers/2603.17187)! Check it out!
- **[03/16/2026]** **v0.3.2** — Multi-claw support: IronClaw, PicoClaw, ZeroClaw, CoPaw, NanoClaw, and NemoClaw now supported alongside OpenClaw. NanoClaw connected via new `/v1/messages` Anthropic-compatible endpoint; NemoClaw via OpenShell inference routing. Added OpenRouter as a supported LLM platform.
- **[03/13/2026]** **v0.3.1** — MinT backend support: RL training now works with both Tinker and MinT. Configurable via `rl.backend` (auto/tinker/mint).
- **[03/13/2026]** **v0.3** — Continual meta-learning support: slow RL updates now only run during sleep hours, idle time, or Google Calendar meetings. Added support/query set separation to prevent stale reward signals from polluting model updates.
- **[03/11/2026]** **v0.2** — One-click deployment via `metaclaw` CLI. Skills enabled by default, RL is now opt-in.
- **[03/09/2026]** We release **MetaClaw** — Just talk to your agent and let it evolve automatically. **NO** GPU deployment required; just plug into the **API**.

---

## 🎥 Demo

https://github.com/user-attachments/assets/d86a41a8-4181-4e3a-af0e-dc453a6b8594

---

## 📖 Overview

**MetaClaw is an agent that meta-learns and evolves in the wild.**
Just talk to your agent as you normally would — MetaClaw turns every live conversation into a learning signal, enabling the agent to continuously improve through real-world deployment rather than offline training alone.

Under the hood, it places your model behind a proxy that intercepts interactions from your personal agent (OpenClaw, CoPaw, IronClaw, PicoClaw, ZeroClaw, NanoClaw, NemoClaw, or any OpenAI-compatible client), injects relevant skills at each turn, and meta-learns from accumulated experience. For Anthropic-native agents like NanoClaw, MetaClaw also exposes a `/v1/messages` Anthropic-compatible endpoint so the full pipeline works without any agent-side changes. Skills are summarized automatically after each session; with RL enabled, a meta-learning scheduler defers weight updates to idle windows so the agent is never interrupted during active use.

No GPU cluster required. MetaClaw works with any OpenAI-compatible LLM API out of the box, and uses a Tinker-compatible backend for cloud-based LoRA training. [Tinker](https://www.thinkingmachines.ai/tinker/) is the default reference path; MinT and Weaver can be enabled through separate compatibility packages when needed.

## 🤖 Key Features

### **One-click deployment**
Configure once with `metaclaw setup`, then `metaclaw start` brings up the proxy, injects skills, and wires your chosen personal agent (OpenClaw, CoPaw, or IronClaw) automatically. No manual shell scripts needed.

### **Three operating modes**

| Mode | Default | What it does |
|------|---------|--------------|
| `skills_only` | | Proxy your LLM API. Skills injected and auto-summarized after each session. No GPU/Tinker required. |
| `rl` | | Skills + RL training (GRPO). Trains immediately when a batch is full. Optional OPD for teacher distillation. |
| `madmax` | ✅ | Skills + RL + smart scheduler. RL weight updates only run during sleep/idle/meeting windows. |

### **Long-term memory**
MetaClaw can persist facts, preferences, and project history across sessions and inject relevant context at each turn — so your agent remembers what you've told it, even weeks later.

### **Asynchronous by design**
Serving, reward modeling, and training are fully decoupled. The agent continues responding while scoring and optimization run in parallel.

---

## 🚀 Quick Start

### 1. Install

**OpenClaw (one-click):** use the [v0.4.0](https://github.com/aiming-lab/MetaClaw/releases/tag/v0.4.0) release—run the snippet below, then `metaclaw setup` and `metaclaw start`. More detail (Windows, mirrors, config, troubleshooting): [`extensions/metaclaw-openclaw/README.md`](./extensions/metaclaw-openclaw/README.md).

```bash
curl -LO https://github.com/aiming-lab/MetaClaw/releases/download/v0.4.0/metaclaw-plugin.zip
unzip metaclaw-plugin.zip -d ~/.openclaw/extensions
openclaw plugins enable metaclaw-openclaw && openclaw gateway restart
```

**pip** (PyPI or this repo):

```bash
pip install -e .                        # skills_only mode (lightweight)
pip install -e ".[rl]"                  # + RL training support (torch, transformers, tinker)
pip install -e ".[evolve]"              # + skill evolution via OpenAI-compatible LLM
pip install -e ".[scheduler]"           # + Google Calendar integration for scheduler
pip install -e ".[rl,evolve,scheduler]" # recomm

[truncated…]