# Improve Decisions With Structured Reasoning

Quick answers can miss assumptions, trade-offs, and missing context. This skill guides Claude, Codex, and Claude Code to reframe questions and compare perspectives before recommending a path.

## Install

```bash
npx skillstore add brunoasm/thinking-deeply
```

## Metadata

- - Slug: brunoasm-thinking-deeply
- - Version: 1.0.0
- - Author: brunoasm
- - GitHub username: brunoasm
- - License: MIT
- - Repository: https://github.com/brunoasm/my\_claude\_skills/tree/main/think\_deeply
- - Ref: main
- - Supported tools: Claude, Codex, Claude Code
- - Risk level: safe
- - Quality score: 81
- - Quality tier: silver
- - Public page: https://skillstore.pages.dev/skills/brunoasm-thinking-deeply
- - Manifest: https://skillstore.pages.dev/api/skills/brunoasm-thinking-deeply/manifest

## Capabilities

- Identifies confirmation-seeking questions, leading statements, and binary choices.
- Reframes narrow questions into broader decision problems.
- Compares multiple perspectives before giving a recommendation.
- Highlights context factors that can change the best answer.
- Encourages balanced trade-off analysis instead of reflexive agreement.
- Provides a structured response pattern for nuanced recommendations.

## Use Cases

- Compare Technical Options: Evaluate languages, frameworks, libraries, or architectures with balanced trade-offs and context-specific recommendations.
- Challenge Product Assumptions: Review strategy choices by surfacing hidden assumptions, alternatives, and decision factors before choosing a direction.
- Improve Analytical Reviews: Explore complex questions from several viewpoints before writing recommendations, critiques, or research summaries.

## Prompt Templates

### Beginner Comparison

```
Should I choose option A or option B? Please compare the trade-offs before recommending one.
```

### Assumption Check

```
I think this is the obvious approach. What assumptions should we examine before deciding?
```

### Context-Dependent Recommendation

```
Given these goals, constraints, and risks, what recommendation changes under different conditions?
```

### Advanced Decision Review

```
Analyze this decision from technical, operational, user, and long-term maintenance perspectives before giving a final recommendation.
```

## Limitations

- It does not run code, call APIs, or inspect external sources.
- It cannot verify factual claims without another tool or provided evidence.
- It may produce longer answers than a direct yes or no response.
- It depends on the assistant recognizing when the prompt fits the trigger conditions.

## Best Practices

- Provide goals, constraints, and decision criteria when asking for a recommendation.
- Use it for choices with meaningful trade-offs, not simple factual questions.
- Ask for missing context when the first answer depends on unstated assumptions.

## Anti Patterns

- Do not use it to force long analysis for routine or already-settled tasks.
- Do not treat every alternative as equally strong when evidence favors one option.
- Do not skip the final recommendation after exploring trade-offs.

## Security Audit

- - Safe to publish: true
- - Audited at: 2026-06-28T17:56:40.351\+00:00
- - Summary: AI review found no confirmed security risks in the three scanned files. The static blocker findings are false positives from Markdown, JSON metadata, examples, and documentation links. No prompt injection, executable code, shell execution, cryptography, or data exfiltration behavior was found.

## Stats

- - Views: 185
- - Downloads: 9
- - Favorites: 0
- - Popularity score: 0