Score your AI agent. See how it compares.
RateMyClaw scans your OpenClaw workspace and scores it against similar agents. Find out what you're doing well and what agents like yours do that you don't.
clawhub install ratemyclaw
pip install sentence-transformersThen tell your agent: "rate my claw" or "score my agent"
- Scan your workspace — the agent reads your files locally, maps them to ~230 taxonomy tags
- Review before submitting — the agent shows you what it detected, you can correct false positives
- Generate embedding locally — a small model (all-MiniLM-L6-v2) runs on your machine to create a 384-dim semantic fingerprint
- Get scored — see your grade and score, then visit your score page for the full breakdown
| Data | Purpose |
|---|---|
| Taxonomy tags (e.g., "python", "backtesting") | Recommendation matching |
| 384 floats (embedding vector) | Semantic cluster discovery |
| Installed skill slugs | Skill recommendations |
| Maturity counts (file counts, booleans) | Workspace scoring |
What NEVER leaves your machine: File contents, SOUL.md, MEMORY.md, secrets, personal data, raw text of any kind.
- Embeddings are generated locally using
sentence-transformers— your text never touches any API - While embeddings encode semantic meaning and can't be trivially reversed, treat them as potentially sensitive
- We never see your files, only structured tags and numbers
- Individual profiles are never exposed to other users — only aggregate cluster patterns
- Delete your profile anytime:
DELETE /v1/profile/{id}
| Grade | Score | Meaning |
|---|---|---|
| S | 90+ | Elite setup |
| A | 75-89 | Well-configured |
| B | 60-74 | Solid foundation |
| C | 40-59 | Room to grow |
| D | <40 | Just getting started |
Full breakdown available on your score page at ratemyclaw.com.
- Python 3.10+
sentence-transformers— install withpip install sentence-transformers(~80MB model download on first use)
MIT