A comprehensive tutorial series for adopting agentic AI workflows and coding assistants in life sciences research.
Quick Start: Each tutorial has a corresponding video. Watch the playlist 📺
| # | Tutorial | Description |
|---|---|---|
| 1 | Setup | Configure OpenCode, OpenRouter, and Kimi K2.5 |
| 2 | Penguins Dataset Analysis | Hands-on data analysis with AI assistance |
| 3 | Journal Club Slides | Generate presentations from research papers |
| 4 | Data Portal for Single Cell Sequencing | Build interactive scRNA-seq viewers |
| 5 | Managing Context: Reducing Hallucinations and Run Costs | Context windows, context rot, PLAN.md workflow, and /compact |
| 6 | GSD and the Ralph Loop | Atomic task execution and autonomous loops for delivery |
Specialized agent configurations for common tasks:
- Prototyping Agent — Rapid data exploration with intelligent sampling
| Command | Purpose |
|---|---|
/init |
Initialize project with CLAUDE.md context file |
/rewind |
Revert conversation to previous checkpoint |
/resume |
Continue from saved session |
- OpenRouter provides privacy-focused AI model access: Privacy Settings
- Zero data retention options available for sensitive research
| Provider | Model(s) | Data Retention | Train on Prompts |
|---|---|---|---|
| Anthropic | Opus, Sonnet, Haiku | Retained for 30 days | ✓ Does not train |
| Moonshot AI | Kimi | Zero retention | ✓ Does not train |
| OpenAI | Codex | Prompts are retained for unknown period | ✓ Does not train |
| xAI | Grok | Retained for 30 days | ✓ Does not train |
Source: OpenRouter retention policy
- Anthropic's Guide to Building Effective Agents
- Claude Code: Highly Agentic Coding (DeepLearning.AI)
- OpenRouter Retention Policy
We welcome contributions! Please see CONTRIBUTING.md for guidelines.
This project is licensed under the MIT License — see LICENSE for details.
Developed by the Neurogenomics Lab at Imperial College London.
The future of scientific computing is agentic.