Scientific Brainstorming

This is for when you need a thought partner to generate research ideas before you have data in hand. It walks through five phases: understanding your context, divergent exploration using techniques like cross-domain analogies and assumption reversal, making connections between ideas, critical evaluation, and synthesis into next steps. The workflow emphasizes conversation over lecture, with prompts designed to challenge assumptions and push beyond safe ideas. It explicitly distinguishes itself from hypothesis generation, which is for when you already have observations. The included reference file covers structured methods like SCAMPER and TRIZ if you want more framework. Works well for identifying research gaps or getting unstuck on methodology.

Install to Claude Code

npx -y skills add k-dense-ai/scientific-agent-skills --skill scientific-brainstorming --agent claude-code