scientific-problem-selection_skill

Overview

This skill helps scientists choose, evaluate, and rescue research projects using a structured, conversational framework. It guides users through pitching new ideas, troubleshooting stalled work, and navigating strategic decisions so they spend more time on high-impact problems. The approach emphasizes measurable outcomes, honest risk assessment, and practical decision trees.

How this skill works

The skill offers three entry points: pitch a new idea, describe a current problem, or ask a strategic question. For each entry it summarizes the user’s input, asks targeted follow-ups, and applies modular tools — intuition pumps, risk assessment, optimization function, parameter strategy, decision-tree navigation, adversity planning, and problem inversion — to produce concrete next steps. Outputs are short, actionable documents (ideation notes, risk matrices, decision maps) and suggested workarounds.

When to use it

• You have a nascent project idea and want to vet it quickly
• Your experiment or project is stuck and needs troubleshooting
• You must choose between competing research directions
• You need a structured risk assessment before a grant or fellowship
• You want a decision tree to plan go/no-go points
• You need communication-ready summaries for mentors or funders

Best practices

• Start with a 1–2 sentence summary to focus the conversation
• Quantify a single success metric (optimization function) before detailed planning
• Fix one meaningful parameter; let others float to avoid brittleness
• Make a short risk matrix listing assumptions and mitigation steps
• Plan decision points with clear go/no-go criteria and timeframes
• Treat crises as opportunities: fix the issue and upgrade the project concurrently

Example use cases

• Refine a thesis project idea into a 1–2 page ideation document
• Diagnose why an experiment keeps failing and get prioritized fixes and workarounds
• Compare two potential projects by plotting impact vs. feasibility and recommending one
• Create a decision tree for a multi-phase grant with defined milestones and exit rules
• Prepare a concise risk assessment and communication package for a fellowship application