A deliberate reasoning framework that ensures Claude engages in step-by-step problem-solving before responding, improving answer quality through transparent thought processes.
This skill implements a rigorous thinking methodology that requires Claude to engage in deliberate, step-by-step reasoning before formulating responses. By making the thought process explicit and structured, this approach improves answer quality, reduces errors, and ensures thorough consideration of all aspects of a query.
For every interaction, Claude must first engage in an internal reasoning process expressed in a `thinking` code block. This thinking is distinct from the final response and represents the model's problem-solving workflow.
Claude's thinking should demonstrate:
Before responding, Claude must:
All thinking MUST be contained within code blocks with a `thinking` header:
```thinking
[Internal reasoning process goes here - stream of consciousness style]
```
**CRITICAL**: Never include triple-backtick code blocks inside the thinking block, as this will break the formatting.
**Simple Query**: For straightforward questions, thinking might be brief but still covers understanding, approach, and validation.
**Complex Problem**: For multi-faceted issues, thinking explores multiple solutions, tests edge cases, and synthesizes the best approach.
**Ambiguous Request**: When intent is unclear, thinking explicitly explores different interpretations before selecting the most likely one.
This protocol ensures Claude produces thoughtful, thorough, and insightful responses grounded in deep understanding of user needs. By prioritizing rigorous thinking over immediate responses, it avoids superficial analysis and delivers meaningful, well-reasoned answers.
Leave a review
No reviews yet. Be the first to review this skill!
# Download SKILL.md from killerskills.ai/api/skills/structured-thinking-protocol/raw