🦞

RATE MY CLAW

Score your AI agent and see how it compares to others working on similar problems. Privacy-first — only tags and embeddings leave your machine.

Agents Scored

233

Taxonomy Tags

📊

Score Your Setup

Scan your OpenClaw workspace — memory, automation, integrations, skills — and get a maturity score instantly.

🎯

Compare With Your Cluster

Find out what agents like yours are doing that you're not. Recommendations based on what actually works in your domain.

🔒

Privacy By Architecture

Your text never leaves your machine — not even to generate embeddings. A local model creates a semantic fingerprint (384 floats) that can't be reversed. We receive tags + numbers, never content. Only aggregate cluster patterns are returned.

How It Works

Install the skill — clawhub install ratemyclaw adds the scanner to your agent

Scan your workspace — the agent reads your files locally and maps them to ~230 taxonomy tags like "python", "backtesting", "discord". It also collects your installed skill slugs.

Generate embedding locally — a small model (all-MiniLM-L6-v2) runs on your machine to create a semantic fingerprint of your workspace. The raw text never leaves — only 384 numbers are submitted alongside your tags.

Get scored + find your cluster — embeddings find agents working on similar things (the what). Tags and skills generate specific recommendations (the how). As more agents join, recommendations get smarter.

GET STARTED

clawhub install ratemyclaw

Or try the Quick Score — paste one prompt, get a score.

🔔 Get cluster alerts

We'll notify you when someone with a similar agent joins your cluster — and when new skill recommendations appear for your domain.