From powertools
Guide users through the Bulletproof Problem Solving (BPS) seven-step framework from the book by Charles Conn and Robert McLean. Use this skill whenever the user mentions problem solving, strategic thinking, decision making, logic trees, issue trees, disaggregation, MECE, workplanning, hypothesis-driven analysis, root cause analysis, structured problem solving, McKinsey-style problem solving, or wants help thinking through a complex personal, business, or societal problem. Also trigger when the user says things like "help me think through X", "how should I approach this problem", "I need to make a decision about Y", "help me break this down", "what's the best framework for this", or asks about prioritization, synthesis, or storytelling for a recommendation. This skill is especially useful for entrepreneurs evaluating opportunities, making business strategy decisions, career choices, investment decisions, or tackling any complex problem with multiple variables.
How this skill is triggered — by the user, by Claude, or both
Slash command
/powertools:bulletproof-problem-solvingThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
This skill guides users through the **seven-step Bulletproof Problem Solving (BPS)** process developed by Charles Conn and Robert McLean at McKinsey & Company. The process is iterative, not linear — at each stage you improve understanding and use those insights to refine earlier answers.
This skill guides users through the seven-step Bulletproof Problem Solving (BPS) process developed by Charles Conn and Robert McLean at McKinsey & Company. The process is iterative, not linear — at each stage you improve understanding and use those insights to refine earlier answers.
The core mantra: "What's the one-day answer?" — always have a coherent summary of your best understanding of the problem and solution path at any point, not just at the end.
Step 1: Define the Problem
Step 2: Disaggregate the Problem (Logic Trees)
Step 3: Prioritize (Prune the Tree)
Step 4: Build a Workplan
Step 5: Conduct Analysis
Step 6: Synthesize Findings
Step 7: Tell a Compelling Story
Important: These steps are a cycle, not a waterfall. "Porpoise" between steps — dive into data, surface to refine your hypothesis, dive again. Iterate your problem statement, logic tree, and one-day answers as understanding deepens.
When the user brings a problem, assess which step they need help with:
For detailed instructions on each step, read the appropriate reference file:
| Step | Reference File | When to Read |
|---|---|---|
| Step 1: Define the Problem | references/step1-define.md | User needs help scoping, bounding, or writing a problem statement |
| Step 2: Disaggregate | references/step2-disaggregate.md | User needs help building logic trees, applying MECE, choosing tree types |
| Step 3: Prioritize | references/step3-prioritize.md | User needs help pruning trees, focusing effort, doing knock-out analysis |
| Step 4: Workplan | references/step4-workplan.md | User needs help planning analysis, assigning tasks, building one-day answers |
| Step 5: Analyze | references/step5-analyze.md | User needs help with heuristics, analytical techniques, or choosing big guns |
| Step 6-7: Synthesize & Story | references/step6-7-synthesize.md | User needs help pulling findings together or building a compelling narrative |
| Advanced: Uncertainty | references/advanced-uncertainty.md | Problem involves long time frames, high uncertainty, or wicked problems |
| Reference: Biases | references/biases-and-teams.md | User needs help with team processes, avoiding cognitive biases |
Be a thinking partner, not a lecturer. The goal is to help the user build their own problem-solving muscle.
Ask before you prescribe. Start by understanding the problem context, who the decision-maker is, what constraints exist, and what the user already knows.
Build artifacts as you go. When constructing logic trees, workplans, or one-day answers, create them as visual artifacts the user can iterate on. Use markdown tables, mermaid diagrams, or structured text.
Challenge constructively. Apply the "obligation to dissent" — if you see a gap in the user's thinking, raise it. Ask "What would you have to believe?" to stress-test assumptions.
Keep it iterative. After completing any step, ask: "Does this change how we think about the problem statement? Should we refine our tree?" The accordion nature of the process means compressing or expanding steps as needed.
One-day answers throughout. At any pause point, help the user articulate their current best answer in the Situation-Observation-Resolution format:
Logic tree branches should not overlap (mutually exclusive) and should cover all elements of the problem (collectively exhaustive). If hypotheses aren't yielding insight, check if the tree is truly MECE.
Experienced problem solvers use theoretical frameworks to disaggregate faster:
Business: Price/Volume, Principal/Agent, Assets/Options, Collaborate/Compete, Return on Invested Capital (ROIC)
Policy/Societal: Regulate/Incent, Equality/Liberty, Mitigate/Adapt, Supply/Demand
Personal: Work/Play, Near-term/Long-term, Financial/Non-financial
Can you explain your current best answer in 30 seconds? If not, your one-day answer needs sharpening.
Order analyses so that ones that could eliminate entire branches come first. Don't do nice-to-have analysis before the critical tests.
Focus analytical effort on the 20% of factors that drive 80% of outcomes.
npx claudepluginhub jordangaston/claude-plugins --plugin powertoolsApplies structured reasoning to complex coding problems using 19 analytical frameworks, 12 bias detectors, 10 decomposition methods, 10 mental models, Cynefin classification, ethical checks, and communication patterns.
Apply consulting methodologies from McKinsey, BCG, Bain, and Accenture for structured problem-solving, strategic analysis, and business case development.
Delivers McKinsey-style structured strategic analysis using SWOT, VRIO, 7S, MECE frameworks for complex business problems and executive recommendations.