CrawlScript

CrawlScript / MMClaw

Public

Ultra-Lightweight, Pure Python Multimodal Agent.

agent agents ai assistant llm-agent
99
7
89% credibility
Found Feb 09, 2026 at 22 stars 5x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

MMClaw is a minimalist pure Python tool for running autonomous AI agents that interact via terminal or messaging apps like Telegram, WhatsApp, and Feishu while using built-in abilities for tasks like command running and file handling.

How It Works

1
🔍 Discover MMClaw

You hear about a super simple AI helper that runs on any computer and chats like a smart friend.

2
📦 Get it on your computer

With one quick action, you add this lightweight tool to your setup—no fuss, just ready to go.

3
⚙️ Personalize your AI buddy

You pick a smart thinking service and choose how to chat, like directly on your screen or through your favorite messaging app.

4
Pick your chatting style
💻
Screen chat

Talk right in the program window, perfect for quick tests.

📱
Messaging apps

Link to Telegram, WhatsApp, or Feishu so you chat from your phone anytime.

5
🚀 Start the conversation

Fire it up, say hello, and watch your AI come alive, ready to listen and help.

6
🛠️ Give it real tasks

Ask it to run quick jobs, read files, or wait patiently— it handles everything smoothly on its own.

🎉 Your personal AI sidekick

Now you have a trusty helper that works across your devices, making life easier every day.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 22 to 99 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is MMClaw?

MMClaw is an ultra-lightweight, pure Python multimodal agent kernel that runs autonomous AI bots via terminal, Telegram, WhatsApp, or Feishu channels. It solves the bloat problem in agent frameworks by ditching Node.js, Docker, and C-extensions for a pip-installable setup that handles tool calls, state management, and LLM interactions out of the box. Users get a customizable engine for building agents that execute shell commands, read/write files, or send media—cross-platform from Raspberry Pi to desktops.

Why is it gaining traction?

It cuts through complex agent stacks with a "batteries-included" yet minimalist core: pure Python means no compile headaches, and CLI commands like `mmclaw run` or `mmclaw config` launch interactive sessions in seconds. Devs love the readable architecture as a live tutorial for agent loops, plus native support for popular chat protocols without extra servers. The hook is extending it effortlessly for custom tools while staying under 100% Python purity.

Who should use this?

Python scripters automating workflows via Telegram or WhatsApp bots, indie devs prototyping edge AI agents on low-spec hardware, or teams auditing agent logic without vendor lock-in. Perfect for solo hackers needing quick file ops or shell integration in personal assistants, not enterprise-scale orchestration.

Verdict

Early alpha with 23 stars and 0.9% credibility score signals low maturity—no tests, sparse examples—but strong README and zero-dep purity make it worth a spin for lightweight experiments. Adopt if simplicity trumps polish; otherwise, wait for more traction.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.