Key Takeaways
- OpenClaw is a personal AI assistant connecting 20+ messaging apps — here's how to set it up for biology research
- Three-step setup: install → configure LLM provider → add bioinformatics skills
- Add K-Dense scientific skills (170+) for domain expertise in RNA-seq, genomics, drug discovery
- Connect to lab messaging: Slack, Telegram, Feishu, Discord, WhatsApp
- Use MCP servers for database access: PubMed, UniProt, NCBI
- For specialized needs, consider MedgeClaw (biomedical) or BioClaw (chat-based bioinfo)
Quick Start
Step 1: Install OpenClaw
npm install -g openclaw@latest
openclaw onboardStep 2: Configure LLM Provider
# In the onboarding wizard, select your provider:
# - Anthropic (Claude) — recommended for research
# - OpenAI (GPT-4)
# - Google (Gemini)
# - Ollama (local, privacy-safe)Step 3: Add Biology Skills
# K-Dense scientific skills (170+ across 17 domains)
git clone https://github.com/k-dense/scientific-skills.git
cp -r scientific-skills/skills/* ~/.openclaw/skills/
# Or just bioinformatics-specific skills
cp -r scientific-skills/skills/rna-seq/* ~/.openclaw/skills/
cp -r scientific-skills/skills/scrna-seq/* ~/.openclaw/skills/Connecting to Lab Messaging
| Platform | Setup |
|---|---|
| Slack | Create bot, add to workspace, configure in OpenClaw |
| Telegram | Create bot via BotFather, add token |
| Feishu | Create app in Feishu developer console |
| Discord | Create bot, add to server |
| Via WhatsApp Web bridge |
When to Use Specialized Forks
| Need | Tool | Why |
|---|---|---|
| General biology research | OpenClaw + K-Dense skills | Maximum flexibility |
| Biomedical research (RNA-seq, drug discovery) | MedgeClaw | 140 skills, research dashboard |
| Quick bioinformatics from chat | BioClaw | 17 skills, any messaging app |
| Edge-cloud with privacy | EdgeClaw | Sensitive data stays local |
| Multi-omics integration | OmicsClaw | 4 omics types, integration workflows |
FAQ
Q1: Do I need coding experience?
Minimal. The onboarding wizard handles configuration. Skills are triggered by natural language. Advanced customization requires some TypeScript knowledge.
Q2: Can I use a local LLM instead of cloud APIs?
Yes. Configure Ollama as your LLM provider for fully local, privacy-safe operation. Performance depends on your hardware and model choice.
Q3: How do I add MCP servers for database access?
Configure MCP servers in ~/.openclaw/config.json:
{
"mcpServers": {
"pubmed": { "command": "npx", "args": ["pubmed-mcp-server"] }
}
}Summary
OpenClaw is the foundation for biology AI assistants in the Claw ecosystem. Install, add K-Dense skills, connect your lab's messaging app, and you have a research assistant accessible from anywhere. For specialized needs, choose from MedgeClaw, BioClaw, EdgeClaw, or OmicsClaw — all built on the same OpenClaw platform.