Reasoning Approach Selection for Claude Prompts

Calibration: Tier 1 (Model-compatible) - runs cleanly on the current dual-primary tier (Opus 5, Sonnet 5) as well as Haiku 4.5 with extended thinking. Structured retrieval, rule evaluation, or template lookup - output shape does not depend on model class. Correct-shape output also on Opus 4.8 (fallback-graceful) and Sonnet 4.6 (legacy-graceful). See repository README for model compatibility.

Select the reasoning approach that matches how the task requires Claude to think. The reasoning layer is the highest-leverage component in a prompt — the difference between shallow and deep output almost always traces back to reasoning quality.

How to Use This Skill

1. Identify the dominant thinking mode the task requires (see the selection guide below).
2. Read the corresponding reference file for that category to find the specific variant that fits.
3. Copy and adapt the reasoning instructions into your prompt.
4. If the task spans categories, see "Combining Reasoning Approaches" below.

Selection Guide

Analytical / Evaluative

Choose when the task requires evaluating, diagnosing, or assessing something that exists.

Approach	Best For	Key Differentiator
General Analysis	Broad evaluation of a situation, opportunity, or problem	Works forward from evidence to conclusions
Root Cause Diagnosis	Something is failing or underperforming — find out why	Works backward from symptoms to causes
Risk Assessment	Evaluating what could go wrong with a decision or plan	Focuses on failure modes, not current evidence

Routing logic: If the task says "evaluate" or "analyze" → General Analysis. If something is broken or underperforming → Root Cause Diagnosis. If the question is "should we do this?" with emphasis on downside → Risk Assessment.

See references/analytical-reasoning.md for complete approaches with usage guidance and failure modes.

Strategic / Planning

Choose when the task requires making decisions about direction, resources, or organizational change.

Approach	Best For	Key Differentiator
Market & Competitive Strategy	Evaluating market opportunities and competitive positioning	Centers on competitive dynamics and differentiation
Resource Allocation	Distributing limited resources across competing priorities	Centers on scarcity and tradeoffs
Change & Transformation	Planning significant organizational or process changes	Centers on transition states and human resistance

Routing logic: If the task involves competitors or market position → Market & Competitive Strategy. If the core tension is "we have X resources and Y demands" → Resource Allocation. If the task involves moving an organization from state A to state B → Change & Transformation.

See references/strategic-reasoning.md for complete approaches with usage guidance and failure modes.

Creative / Generative

Choose when the task requires generating something new rather than analyzing something that exists.

Approach	Best For	Key Differentiator
Concept Development	Generating a new concept, design, or creative direction	Explores the possibility space, then develops the best direction
Messaging & Narrative	Crafting messages, stories, or positioning	Driven by audience psychology and narrative arc
Solution Ideation	Creative problem-solving for a defined problem	Problem is known; solution space is open

Routing logic: If the deliverable is a creative concept or design → Concept Development. If the deliverable is a message, story, or narrative → Messaging & Narrative. If there's a defined problem needing creative solutions → Solution Ideation.

See references/creative-reasoning.md for complete approaches with usage guidance and failure modes.

Technical / Problem-Solving

Choose when the task involves building, fixing, or migrating technical systems.

Approach	Best For	Key Differentiator
System Design	Designing a new system or architecture from requirements	Building something new — key decisions and tradeoffs
Debugging & Incident Analysis	Finding and fixing something broken	Hypothesis-driven troubleshooting of existing systems
Migration & Transition	Moving from one system or platform to another	Managing the transition while maintaining operations

Routing logic: If building something new → System Design. If something is broken → Debugging & Incident Analysis. If moving between systems → Migration & Transition.

See references/technical-reasoning.md for complete approaches with usage guidance and failure modes.

Research / Synthesis

Choose when the task requires processing multiple information sources into coherent analysis.

Approach	Best For	Key Differentiator
Evidence Synthesis	Integrating multiple sources into evidence-grounded conclusions	Organizes by theme, discriminates by source quality
Landscape Scan	Broad overview of a domain — players, trends, maturity	Optimizes for breadth and orientation, not depth
Gap Analysis	Comparing current state vs. desired state	Explicitly comparative framing — what exists vs. what should

Routing logic: If the task is "what does the evidence say about X?" → Evidence Synthesis. If the task is "what's out there in domain X?" → Landscape Scan. If the task is "where do we fall short of target?" → Gap Analysis.

See references/research-reasoning.md for complete approaches with usage guidance and failure modes.

Comparative / Decision-Support

Choose when the task requires comparing options and making a selection or ranking.

Approach	Best For	Key Differentiator
Option Evaluation	Comparing multiple options against criteria to select one	General-purpose comparison with decisive recommendation
Vendor / Tool Selection	Comparing products, vendors, or platforms for adoption	Adds total cost of ownership and integration concerns
Prioritization	Ranking a set of items by priority	Produces tiers and sequences, not a single winner

Routing logic: If selecting one option from several → Option Evaluation. If the options are products, tools, or vendors → Vendor / Tool Selection. If ranking a list rather than picking a winner → Prioritization.

See references/comparative-reasoning.md for complete approaches with usage guidance and failure modes.

Combining Reasoning Approaches

Some tasks span multiple categories. When combining, follow these principles:

Keep the total steps to 5-7. More than 7 steps and Claude treats each one less carefully. Combine steps from different approaches rather than concatenating entire approaches.

Lead with the dominant task type. If it's primarily a strategic decision that needs some technical evaluation, use the strategic reasoning structure and fold in the technical steps — not the other way around.

Watch for contradictions. Different approaches sometimes push in opposite directions (e.g., creative approaches say "explore broadly" while analytical approaches say "narrow to the core question"). When combining, make the sequence explicit: "First, explore broadly. Then, evaluate the most promising directions."

Example: Strategic Decision with Technical Evaluation

<reasoning>
1. Define the strategic objective and the constraints that bound the decision.
2. Generate 3 structurally different approaches to achieving the objective.
3. For each approach, evaluate the technical feasibility: can our team build this with available resources and timeline? What are the technical risks?
4. Assess each approach against the strategic criteria: competitive positioning, resource efficiency, and alignment with stated priorities.
5. Identify second-order consequences — what does each approach enable or prevent later?
6. Recommend the approach with the best combined strategic-technical tradeoff profile.
</reasoning>

Common Combinations

Task Shape	Lead With	Fold In
Strategic decision needing technical vetting	Strategic	Technical (feasibility check)
Creative concept needing market validation	Creative	Strategic (competitive positioning)
Technical design needing business justification	Technical	Analytical (risk + value assessment)
Research synthesis needing actionable priorities	Research	Comparative (prioritization)
Risk assessment needing root cause depth	Analytical (risk)	Analytical (root cause)

Critical: Output Quality

Match reasoning depth to task complexity. A simple comparison needs 4-5 steps, not a 7-step methodology. When the prompt produces shallow output, the fix is usually one of these:

• Wrong category: The reasoning approach doesn't match the actual thinking required. Re-read the routing logic above.
• Too many steps: Claude spreads attention thin. Cut to the 5 most important steps.
• Missing context: The reasoning approach is right, but Claude lacks the domain information to apply it well. The fix is in the context layer of the prompt, not the reasoning layer.
• Generic output: Claude is applying the steps mechanically rather than to the specific situation. Add specificity to the prompt's context and identity layers, not more reasoning steps.

Domain Specialization

The 18 approaches in this Skill are domain-agnostic — they work across contexts. For tasks requiring deeper domain specialization, domain-specific reasoning approaches are available in the rootnode-domain Skills (business strategy, software engineering, content & communications, research & analysis, agentic & context engineering) if installed. Domain approaches extend these core approaches — they do not replace them.

Troubleshooting

Output is shallow despite using a reasoning approach: Check whether the approach category matches the actual thinking required. A strategic task using an analytical approach will produce analysis when you need decisions.

Claude ignores some reasoning steps: Too many steps. Cut to 5-7 maximum. Claude gives diminishing attention to steps beyond 7.

All outputs from this approach look the same: The reasoning is right but the context is insufficient. Claude needs specific details about the situation to produce specific analysis. Add concrete context to the prompt.

Claude applies a standard framework instead of thinking specifically: Some approaches (especially strategic ones) trigger Claude's tendency to apply textbook frameworks (Porter's Five Forces, SWOT, etc.) rather than analyzing the specific situation. Add: "Analyze the specific dynamics of this situation. Do not apply standard frameworks unless they genuinely illuminate the problem."