🎓 Exam2Knowledge
_v1.0.0

中文 | English

"See question → Recognize pattern → Apply method → Get answer"

---

✨ What It Does

Transforms exam questions into structured, exam-focused knowledge systems. Not systematic learning — this builds a pattern-recognition reflex for instant problem-solving.

Perfect for: Past exam papers, practice questions, course materials → High-frequency test points → Strategic study plan

📁 Structure

.agents/skills/Exam2Knowledge/
├── SKILL.md                  # Core definition + processing pipeline
├── assets/
│   └── output-template.md    # Standardized output structure
└── references/
    ├── blooms-taxonomy.md    # 6-level cognitive hierarchy
    ├── diagnostic-rubric.md  # 5-category error classification
    └── quantified-rubric.md  # Quantified self-check standards

⚙️ How It Works

1. Pattern Recognition — Identify question type & cognitive level (Bloom's Taxonomy)
2. Reverse Engineering — Extract & tier knowledge points ([**] Must-Master to [] Optional)
3. Solution Template — Build universal steps + decision rules
4. Error Pattern Library — Catalog mistakes by category (Diagnostic Rubric)
5. Cross-Question Intelligence — Frequency ranking + clusters + trends (batch mode; self-check standards in Quantified Rubric)
6. Exam-Ready Output — Prioritized action framework (Output Template)

🎯 Decision Rules

Scenario	Action
Single question	Full 5-step analysis + solution template
Multiple questions (≥5)	+ frequency statistics + clusters
Weak spot diagnosis	Strengthen error pattern analysis
Image/OCR input	Verify OCR accuracy first
Humanities subject	Switch to humanities analysis mode

✨ Key Features

• 🎯 5-step processing pipeline with strict execution order
• 🔀 Conditional logic — output adapts to input type & time constraint
• ✅ Built-in self-check — format, content, consistency, boundary verification
• 📚 Tiered knowledge — [] / [] / [] frequency-based ranking
• 🛡️ Capability boundaries — explicit "can do / cannot do" declaration

📥 Supported Input

• ✅ Single/multiple questions or full exam papers (text format)
• ✅ OCR results from handwritten exams
• ✅ All subjects: Math, Physics, Chemistry, Biology, History, Languages...
• ❌ Raw image input (require OCR text first)

📄 License

MIT

---