meow:sequential-thinking

Structured step-by-step reasoning with hypothesis generation, elimination, revision, and deterministic thought tracking.

What This Skill Does

meow:sequential-thinking prevents the "guess root causes" anti-pattern. Instead of jumping to "I think it's X", it enforces evidence-based reasoning: generate hypotheses from observations, test each against evidence, eliminate until root cause is confirmed.

Called by meow:fix during diagnosis, but also useful standalone for complex architecture decisions or multi-step analysis.

Core Capabilities

• Hypothesis-driven investigation — generate → test → eliminate → conclude
• Evidence-based only — every hypothesis needs verifiable evidence
• Revision capability — explicitly revise when new evidence contradicts
• Branching — explore 2-3 alternatives, converge with decision rationale
• Structured output — hypothesis table + elimination + root cause conclusion
• Deterministic scripts — process-thought.js tracks history/branches, format-thought.js outputs box/markdown/JSON
• Advanced techniques — spiral refinement, multi-branch convergence, progressive context deepening
• Concrete examples — debugging, API design, architecture decision walkthroughs

When to Use This

Use meow:sequential-thinking when...

• Root cause isn't obvious (multiple possibilities)
• meow:fix invokes diagnosis phase
• Architecture decision with competing approaches
• Any "I think it's X" needs evidence before acting

Scripts

Script	Purpose	Usage
process-thought.js	Validate, track history, branches, revisions (max 20)	node scripts/process-thought.js --thought "..." --number 1 --total 5 --next true
format-thought.js	Format as box/simple/markdown/JSON	node scripts/format-thought.js --thought "..." --number 1 --total 5 --format json
test-scripts.sh	13-test suite for both scripts	sh scripts/test-scripts.sh

The --summary flag on process-thought.js produces a context-efficient handoff for meow:fix.

References

Reference	When to load
hypothesis-testing.md	Always — output template for diagnosis
core-patterns.md	When using revision or branching
advanced-techniques.md	Complex problems (spiral refinement, multi-branch convergence)
advanced-strategies.md	Uncertainty or revision cascades
examples-debug.md	Performance debugging walkthrough
examples-api.md	API design reasoning
examples-architecture.md	Architecture decision
gotchas.md	Common reasoning mistakes

Skill Details

Phase: Called by meow:fix (Step 2 Diagnose). Also standalone.

Gotchas

• Premature conclusion: test minimum 2 hypotheses before concluding
• Evidence-free hypotheses: every hypothesis needs verifiable evidence
• Confirmation bias: actively search for evidence AGAINST each hypothesis

Related

• meow:fix — Calls sequential-thinking during diagnosis
• meow:investigate — Collects evidence that feeds into sequential-thinking