Smithery is now a part of Arcade.dev! Read more in our announcement here
Three-pass critical reading framework for systematic document analysis. Supports tech blogs, retrospectives, technical documentation, personal writing, and academic papers...
A systematic approach to deep document analysis based on the three-pass reading methodology. This skill guides critical analysis with emphasis on Third Pass: assumption validation, context dependencies, and actionable reconstruction.
This skill provides structured guidance for Third Pass critical analysis of documents. Users complete First and Second Pass independently (reading, understanding, note-taking), then use this skill for deep critical analysis that challenges assumptions, identifies context dependencies, and extracts actionable insights.
Typical workflow:
Automatically detect and adapt analysis to document type based on content, structure, vocabulary, and writing style.
Supported types:
Detection approach:
Follow these steps systematically:
State the detected type and confirm with user: "이 문서를 [TYPE]로 판단했습니다. Third Pass 비판적 분석을 시작할까요?"
Challenge core claims and validate reasoning.
Universal questions (all types):
Load type-specific questions: Read references/critical-questions.md for detailed questions per document type.
Output format:
#### Critical Analysis
- **주장의 약점 / 적용 한계**
- [Specific issues with claims, evidence, or reasoning]
- [Scenarios where this approach fails]
- [Generalization limits]
Identify implicit assumptions that constrain validity.
Key dimensions:
Output format:
- **컨텍스트 의존성 (환경, 스택, 시기)**
- Environment: [Tech stack, scale, constraints that this assumes]
- Prerequisites: [What reader needs to know/have]
- Temporal: [Time-dependent assumptions]
- Cultural: [Organizational or regional assumptions]
Find gaps, alternatives, and unexplored angles.
Look for:
Output format:
- **누락된 관점 / 대안**
- Alternatives: [Other approaches not considered]
- Trade-offs: [Costs or sacrifices not discussed]
- Contrary views: [Opposing perspectives]
- Edge cases: [Scenarios not addressed]
Demonstrate deep understanding by restating core ideas without reference to source.
Tasks:
Output format:
#### Reconstruction
- **핵심을 나만의 언어로**
- [Restate the central thesis and key supporting ideas without looking at source]
- **내 상황에 적용한다면**
- 직접 적용: [What can be adopted as-is]
- 변형 필요: [What needs modification and how]
- 적용 불가: [What doesn't fit my situation and why]
Extract practical value and next steps.
Focus on:
Output format:
#### Discussion
- **가져갈 것**
- [Most valuable insights - what to remember and apply]
- **한계**
- [Boundaries of applicability, remaining uncertainties]
#### Action Items
- [ ] [Specific action item or research question]
- [ ] [Related concept to explore]
#### Future Ideas
- [ ] [Experiment to validate understanding]
Guide user through analysis with checkpoints:
After type detection: "[TYPE]로 분석을 시작합니다. 이 유형에 맞는 비판적 질문을 적용하겠습니다."
After each major step: "[STEP] 완료했습니다. 다음 단계([NEXT_STEP])로 진행할까요, 아니면 이 부분을 더 깊이 탐구할까요?"
When user seems uncertain: "특정 섹션이나 주장에 대해 더 자세히 분석하고 싶으신 부분이 있나요?"
Provide targeted assistance beyond Third Pass:
Second Pass support:
Contextual research:
Concept clarification:
Comparison:
For optimal results:
When to use this skill:
✅ After reading complex technical content ✅ When evaluating whether to adopt an approach ✅ Before citing or building upon an article ✅ When reviewing your own writing for gaps ✅ During literature review or research
When NOT to use:
❌ For quick skimming or surface reading ❌ When basic comprehension is sufficient ❌ For simple reference lookups
For complex analysis needs:
references/critical-questions.md - Type-specific analytical questionsreferences/reconstruction-guide.md - Detailed reconstruction techniquesreferences/common-patterns.md - Frequently encountered patterns and pitfalls