Key Takeaways
- 5 medical AI agents share similar names in the OpenClaw ecosystem — this guide disambiguates them
- MedgeClaw (~1,000 stars) is the most mature: Claude Code + 140 K-Dense skills, research dashboard, Docker-based
- MedClaw by zteyesreal (~45 stars): clinical decision support agent
- MedClaw by Paulzhang (~5 stars): Chinese-language medical OpenClaw fork
- MedClaw by Suzano (~0 stars): multi-domain agent (medical, legal, finance)
- MolClaw (~13 stars): molecular science orchestrator, NanoClaw fork
Full Comparison
| MedgeClaw | MedClaw (zteyesreal) | MedClaw (Paulzhang) | MedClaw (Suzano) | MolClaw | |
|---|---|---|---|---|---|
| Stars | ~1,000 | ~45 | ~5 | ~0 | ~13 |
| Focus | Biomedical research | Clinical decision support | Medical-domain OpenClaw | Multi-domain | Molecular science |
| Base | Claude Code + K-Dense | Agent-based clinical AI | OpenClaw fork | Independent | NanoClaw fork |
| Skills | 140 K-Dense scientific | Early stage | Chinese-language | TypeScript multi-domain | OpenAI/OpenRouter |
| Status | Active (weekly updates) | Active | Active | Early stage | Active |
MedgeClaw — The Research Powerhouse
GitHub: xjtulyc/MedgeClaw · ~1,000 ★
Core attributes:
- Architecture: Claude Code + 140 K-Dense scientific skills
- Unique features: Real-time research dashboard, built-in RStudio (8787) and JupyterLab (8888), Feishu rich card output
- Deployment: Docker-based, all computation in isolated containers
- Best for: Serious biomedical research — RNA-seq, drug discovery, clinical data analysis
How to Choose
| If you need... | Choose... |
|---|---|
| Full biomedical research platform | MedgeClaw |
| Clinical decision support | MedClaw (zteyesreal) |
| Chinese-language medical AI | MedClaw (Paulzhang) |
| Multi-domain agent (medical + legal) | MedClaw (Suzano) |
| Molecular science orchestration | MolClaw |
FAQ
Q1: Why are there so many projects with similar names?
Independent teams in different countries chose similar names without coordination. This is a common pattern in the OpenClaw ecosystem — see also ScienceClaw (4 variants) and PaperClaw (6 variants).
Q2: Which one should a biomedical PhD student choose?
MedgeClaw. It has the most skills (140), active development, Docker deployment, and built-in research tools (RStudio, JupyterLab).
Q3: Can I use MedgeClaw skills with other agents?
Yes. K-Dense skills are standard SKILL.md files compatible with any agent supporting the format.
Summary
5 medical AI agents share similar names in the OpenClaw ecosystem. MedgeClaw is the clear leader with 1,000 stars, 140 skills, and active weekly development. The others serve niche use cases. Check the comparison table to find the right fit for your research.