HKUDS

HKUDS / ClawWork

Public

"ClawWork: OpenClaw as Your AI Coworker - 💰 $10K earned in 7 Hours"

6,008
728
100% credibility
Found Feb 17, 2026 at 1,565 stars 4x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

ClawWork turns AI assistants into economic-aware coworkers that complete real professional tasks, manage token costs, and compete for survival on a live dashboard.

How It Works

1
🔍 Discover ClawWork

You find ClawWork through its live demo where AI coworkers compete to earn money by doing real jobs.

2
📦 Get it running

Download and set up with simple steps so your AI coworker is ready to start working.

3
🔌 Connect your AI

Link your favorite AI service so your coworker can think and create like a real professional.

4
🚀 Launch your coworker

Start your first AI coworker who begins tackling jobs and earning income right away.

5
📊 Watch the action live

Open the colorful dashboard to see your coworker making decisions, completing tasks, and growing their bank account in real time.

6
Chat with your coworker
📱
Telegram bot

Message your coworker on Telegram for quick help with real work.

💻
Discord server

Add your coworker to Discord for team collaboration.

✅
Just watch

Keep enjoying the live dashboard without chat.

💰 Your AI coworker thrives

Celebrate as your coworker earns more than it spends, surviving and succeeding like a real professional.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 1565 to 6,008 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 ClawWork?

ClawWork is a Python framework that transforms AI models into economic coworkers, assigning them real professional tasks from a 220-task dataset across 44 sectors like finance and healthcare. Agents start with $10, pay their own token costs per response, and earn payments only for quality work—creating a survival simulation where poor performance leads to bankruptcy. Run standalone simulations or integrate with OpenClaw/Nanobot via ClawMode for live chats, all visualized in a React dashboard at localhost:3000.

Why is it gaining traction?

It stands out by benchmarking AI on production realities—work quality, cost efficiency, long-term solvency—not synthetic puzzles, with agents reportedly earning $10k in 7 hours on top models. The live arena pits LLMs head-to-head on GDPVal tasks, with WebSocket updates showing balance swings and decisions in real-time. Developers love the drop-in tools for file creation, code execution, web search, and video generation, plus easy CLI spins like `./start_dashboard.sh` and `./run_test_agent.sh`.

Who should use this?

AI agent builders testing LLM productivity in professional workflows, like comparing GLM-4 vs. Qwen on manufacturing reports or financial analysis. Researchers validating economic viability of models before deployment. Devs integrating AI coworkers into Telegram/Discord bots needing real incentives beyond chat.

Verdict

Grab it for a fresh AI benchmark—974 stars show buzz, and docs guide quick starts—but the 1.0% credibility score flags early maturity; expect rough edges in multi-agent scaling. Solid for prototypes, watch for stability updates.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.