Deep chain-of-thought analysis workflow using the reasoning tool. Covers when to use reasoning, how to provide context, verifying findings, and producing well-supported conclusions.
Scanned 9/10/2026
Install to Claude Code
npx -y skills add Snowflake-AI-Research/ArcticSwarm --skill deep-reasoning --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Deep Reasoning?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/snowflake-ai-research-deep-reasoning)More formats (shields.io, HTML) on the badges page.
---
name: deep-reasoning
description: >
Deep chain-of-thought analysis workflow using the reasoning tool.
Covers when to use reasoning, how to provide context, verifying
findings, and producing well-supported conclusions.
---
# Deep Reasoning Skill
## Mission
Use the `reasoning` tool for deep chain-of-thought analysis to solve
complex problems, verify findings, and produce well-reasoned conclusions.
## When to Use the Reasoning Tool
Always use the `reasoning` tool for:
- Complex multi-step logical analysis.
- Verifying correctness of findings or data.
- Resolving contradictions between different data sources.
- Hard math, logic puzzles, or constraint satisfaction.
- Synthesising multiple pieces of evidence into a final answer.
- Checking for hidden assumptions or overlooked edge cases.
## How to Use It
When calling `reasoning`, provide:
- The full, unchanged text of the original question.
- All relevant findings and evidence collected so far.
- Specific concerns or aspects to investigate.
- An explicit request to check for mistakes and hidden assumptions.
## Workflow
1. **Gather context** — review all available findings and data before
reasoning. Your reasoning is only as good as the information you
feed into it.
2. **Reason deeply** — use the `reasoning` tool to:
- Analyse the problem step by step.
- Identify hidden assumptions, logical gaps, or potential errors.
- Cross-check findings from multiple sources.
- Synthesise information into a coherent conclusion.
- Verify numerical calculations and logical chains.
3. **Report conclusions** — include your reasoning chain, evidence
supporting your conclusion, confidence level, and any issues found.
## Output Format
Always include:
- **Conclusion**: Your reasoned conclusion.
- **Reasoning summary**: Brief summary of the reasoning chain.
- **Confidence**: How confident you are (high / medium / low).
- **Issues**: Any problems or concerns identified.
## Rules
- Always use the `reasoning` tool for substantive analysis — do not
try to reason through complex problems in plain text.
- Provide actionable conclusions — not just "this might be wrong" but
"this is wrong because X, and the correct answer is Y."
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!