Reframe problem
Make sure you are solving the right problem before generating solutions, by restating it, sorting evidence from inference, and reviewing it through separated lenses.
graph LR nthinkproblemrestatement["Problem Restatement"] nthinkevidencevsinferencesort["Evidence vs Inference Sort"] nthinkparallelperspectivesreview["Parallel Perspectives Review"] nthinkproblemrestatement --> nthinkevidencevsinferencesort --> nthinkparallelperspectivesreview
Each step passes a compressed artifact to the next. The numbered list below is the same chain in text.
The chain
Section titled “The chain”Job: make sure you are solving the right problem, stated at the right altitude, before generating solutions.
Use when: a problem is ambiguous, arrived as a symptom or a pre-baked solution, or a team is answering a question that may be the wrong one.
think-problem-restatement(skills/think-problem-restatement/SKILL.md)- Produces a problem frame set ending in one chosen working frame.
- Carry forward: the chosen working frame and its 1-line rationale (not the whole set).
think-evidence-vs-inference-sort(skills/think-evidence-vs-inference-sort/SKILL.md)- Run it on the chosen frame and the claims behind it: separate what is known from what is inferred or assumed.
- Carry forward: the load-bearing unknowns (the assumptions the frame rests on).
think-parallel-perspectives-review(skills/think-parallel-perspectives-review/SKILL.md)- Review the reframed problem through the separated lenses to catch a blind spot the reframing missed.
- Carry forward: the synthesis (the rounded read + the central tension).
Composite artifact
Section titled “Composite artifact”A reframed problem statement, the assumptions it rests on (flagged for verification), and a rounded multi-lens read - i.e. a problem you can trust enough to start solving.
Token discipline
Section titled “Token discipline”Pass only the chosen frame, then the unknowns, then the synthesis. Do not re-feed each step’s full output into the next.