Cognition · Data artifact
Structured Reasoning Prompt Template
Data artifactCognitionCognition & MemoryVariation point (abstract)arc:StructuredReasoningPromptTemplate
A prompt template that prescribes the format in which an agent exposes its reasoning, creating consistent step boundaries and evaluation checkpoints in reasoning traces.
Responsibility. Structures agent reasoning output into a consistent, parseable format.
Also known as: Structured CoT prompt, Structured reasoning template
Variants
| Variant | When to choose |
|---|---|
| Numbered Step Reasoning Template | Choose when the simplest structure suffices and clear step boundaries are needed for automated parsing and step-by-step evaluation. |
| Tagged Reasoning Template | Choose when evaluation must extract and assess specific reasoning components, e.g., whether verification actually checks results or conclusions quantify confidence. |
Relationships
configures structural
Design guidance
- MUST apply whichever reasoning structure is adopted consistently across all reasoning tasks so evaluation pipelines can reliably parse traces.
- SHOULD specify domain-specific steps with concrete deliverables rather than generic 'think step by step' instructions.
- MAY embed per-step verification criteria so agents mark assumptions and revise failing steps before continuing.
Classification
- Patterns
- Structured Chain-of-Thought promptingExplicit criteria checking embedded in reasoning promptDomain-specific reasoning steps (e.g., differential diagnosis, design-before-implementation)
- Quality attributes
- Reliability (ISO/IEC 25010 | NIST AI RMF: valid and reliable)Explainability (NIST AI RMF: explainable and interpretable)
- Risks mitigated
- Naive Chain-of-Thought assumptionVerbose pseudo-reasoning
Sources
- Ch3.9: T. Nguyen, "Reasoning Quality," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 3.9. ISBN: 9798244538229.