Reasoning frameworks and problem decomposition techniques. Use when planning implementation, evaluating arguments, estimating scope, decomposing complex tasks, or applying first principles thinking.
Foundation: Based on OODA Loop's Orient phase (Boyd) - the "cognitive engine" that drives decision-making through mental models, prior experience, and analysis/synthesis.
When to use: Problem framing, solution design, risk assessment, logical validation.
MECE relationship:
@smith-clarity/SKILL.md - Defensive thinking (what to avoid)@smith-validation/SKILL.md - Proving/testing (verifying correctness)General principles to specific conclusions (logically certain):
Pattern: If all A are B, and C is A, then C is B.
Specific observations to general patterns (probabilistic):
Pattern: Observed A1, A2, A3... all have property B. Therefore all A probably have B.
Caution: Inductive conclusions can be wrong.
Best explanation from incomplete observations (inference to best explanation):
Pattern: B is observed. A would explain B. Therefore A is probably true.
Modern LLMs have built-in extended thinking for complex problem-solving.
When to use: Complex architectural decisions, multi-step refactoring, security analysis, performance optimization.
Break down to fundamentals, reason up:
Universally applicable problem-solving:
Understand the problem
Devise a plan
Carry out the plan
Look back
Order-of-magnitude approximation with limited data:
Systematic bottleneck elimination (Goldratt):
Handle uncertainty with optimistic, likely, pessimistic:
Trace symptoms to root cause:
Before implementation, imagine failure has occurred:
Increases problem identification by 30%.
Think backward to avoid failure:
Ensure coverage by examining from 6 perspectives:
When reasoning or planning:
@smith-clarity/SKILL.md - Cognitive guards, logic fallacies@smith-validation/SKILL.md - Hypothesis testing, debugging