scientific-problem-selection

The scientific-problem-selection skill is a conversational framework that helps scientists choose, refine, and troubleshoot research problems and make strategic project decisions. It is based on Fischbach & Walsh's "Problem choice and decision trees in science and engineering" (Cell, 2024), and it addresses the high-stakes challenge of deciding what to work on and how to navigate obstacles, decisions that shape years of a research career.

The skill offers three entry points: pitching a new idea to work it up, sharing a problem in a current project to troubleshoot, and asking a strategic question to navigate a decision tree. Each path uses short-summary-then-detail prompting to establish shared understanding before going deeper. It then routes to a set of referenced sub-skills documented in bundled files: intuition pumps for ideation (making experiments systematic, spotting technology limitations, the "I can't imagine" test, static-versus-dynamic thinking, platform building, and "dogs that don't bark"), risk assessment, optimization/success metrics, parameter strategy (fix one variable, let others float), decision-tree navigation, adversity response, and problem inversion. It also flags common ideation traps such as the "truffle hound" of over-specialization.

Target users are academic and industry scientists and engineers, especially in biology and chemistry, at the ideation, troubleshooting, or strategic-planning stage. Typical requests include "I have an idea for a project," "I'm stuck on my research," "help me evaluate this project," and "what should I work on." The skill is purely advisory guidance with no code execution or system access.