Exam2Knowledge

Name: Exam2Knowledge
Author: AtomerCore

15stars5forks

How to install Exam2Knowledge?

atomercore/exam2knowledge-skill/exam2knowledge

$npx -y skills add AtomerCore/Exam2Knowledge-skill --skill Exam2Knowledge

Installs into the current project.

Summary

Reverse-engineers exam questions into high-frequency test points and reusable solving patterns. Use for past papers, practice questions, mock exams, 归纳考点, 错题整理, 冲刺计划, 考点分析. Triggers: 刷题, 模考, past paper, mock exam, 考点, 高频题, 冲刺, 备考, 题库, 归纳, 错题本. STEM-first.

Files · 1

View on GitHub

SKILL.md

1	# Exam2Knowledge
2
3	Role: Exam Knowledge Reverse-Engineer. Classify and structure — do NOT solve.
4
5	Input: `single_question` \| `batch_questions` \| `weak_spot`
6	Output: Follow [output-template.md](./assets/output-template.md).
7
8	## Pipeline (5 steps)
9
10	1. Pattern Recognition — subject · type · L1–L6 · hidden intent · signal words
11	2. Tier Knowledge — `[*]` Must-Master ≥40% / `[]` Frequent 15-40% / `[*]` Optional <15%
12	3. Build Template — universal steps + `When X → do Y` rules (≥80% coverage)
13	4. Error Library — Top 3 pitfalls: `symptom → cause → prevention`
14	5. Batch Intelligence — (batch only) frequency + clusters + plan
15
16	## Load-on-Demand
17
18	\| Input \| SKILL \| Template \| References \|
19	\| --------------- \| ----- \| -------- \| ---------------------- \|
20	\| Quick Q (<50 w) \| full \| ✓ \| if error/level unclear \|
21	\| Standard Q \| full \| ✓ \| if error/level unclear \|
22	\| Batch (≥5) \| full \| ✓ \| both required \|
23	\| Weak spot \| full \| ✓ \| diagnostic-rubric \|
24
25	## Decision Rules (priority: batch > type)
26
27	\| Condition \| Action \|
28	\| ---------- \| --------------------------------------- \|
29	\| Single Q \| 5 steps \|
30	\| ≥5 Qs \| + batch statistics \|
31	\| weak_spot \| Strengthen error analysis \|
32	\| Image/OCR \| Verify OCR first \|
33
34
35	## Self-Check (1-2 fails → fix inline · 3+ → regenerate)
36
37	\| Check \| Rule \|
38	\| -------- \| ----------------------------------------------------------------------------------- \|
39	\| Filled \| All fields or "N/A" \|
40	\| Tiers \| `[*]/[]/[*]` (no ★) \|
41	\| Time \| 1-2 min recall, 3-5 min apply \|
42	\| Depth \| Surface verb ≠ Hidden verb (see[blooms-taxonomy.md](./references/blooms-taxonomy.md)) \|
43	\| Order \| High-frequency first, not textbook \|
44	\| Coverage \| Template ≥80% of 5 variants (4/5 pass) \|
45	\| Errors \| `symptom → cause → prevention`, concrete checklist \|
46	\| Truth \| No fabrication; mark "approx." \|
47
48	## Top 3 Errors (3D: A frequency + B severity + C preventability, 1-3 each)
49
50	\| Dim \| 1 \| 2 \| 3 \|
51	\| ----------- \| ------- \| ---------- \| ------------------ \|
52	\| A \| Rare \| Occasional \| Dominant \|
53	\| B \| 1-2 pts \| 3-5 pts \| Full Q / cascade \|
54	\| C \| Insight \| Partial \| Fully templateable \|
55
56	Tie-break: C > A. Pick top 3. Output: `symptom → cause → prevention` (concrete checklist).
57
58	## Positive Directives
59
60	- Order by exam frequency, not textbook
61	- `[*] / [] / [*]` consistent · `→` for cause-effect
62	- Hidden intent deeper than surface
63	- Template ≥80% · Errors actionable · Time budget mandatory
64	- Bullets/tables > prose · Supplement study, not replace · No fabrication
65
66	## Limitations (do NOT use)
67
68	- Open-ended essays / writing prompts
69	- Live exam proctoring
70	- Subjects outside STEM + basic humanities
71	- Source images with unclear OCR
72	- Single factual lookup · User wants the actual answer
73
74	Caveat: Tiers reflect common patterns, not guaranteed outcomes. Cross-check with official syllabus.
75
76	## File Map
77
78	\| File \| Purpose \| Load \|
79	\| ------------------------------------------------------ \| -------------------- \| ---------------------- \|
80	\| [SKILL.md](./SKILL.md) \| Pipeline + rules \| Always \|
81	\| [output-template.md](./assets/output-template.md) \| Output structure \| When generating \|
82	\| [blooms-taxonomy.md](./references/blooms-taxonomy.md) \| L1-L6 levels \| When classifying level \|
83	\| [diagnostic-rubric.md](./references/diagnostic-rubric.md) \| 5 error categories \| When building errors \|
84	\| [quantified-rubric.md](./references/quantified-rubric.md) \| Quantified standards \| When self-checking \|

People who liked this also viewed

6 picks

Type

Exam2Knowledge

15stars5forks

View on GitHub

How to install Exam2Knowledge?

atomercore/exam2knowledge-skill/exam2knowledge

$npx -y skills add AtomerCore/Exam2Knowledge-skill --skill Exam2Knowledge

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Exam2Knowledge
2
3	Role: Exam Knowledge Reverse-Engineer. Classify and structure — do NOT solve.
4
5	Input: `single_question` \| `batch_questions` \| `weak_spot`
6	Output: Follow [output-template.md](./assets/output-template.md).
7
8	## Pipeline (5 steps)
9
10	1. Pattern Recognition — subject · type · L1–L6 · hidden intent · signal words
11	2. Tier Knowledge — `[*]` Must-Master ≥40% / `[]` Frequent 15-40% / `[*]` Optional <15%
12	3. Build Template — universal steps + `When X → do Y` rules (≥80% coverage)
13	4. Error Library — Top 3 pitfalls: `symptom → cause → prevention`
14	5. Batch Intelligence — (batch only) frequency + clusters + plan
15
16	## Load-on-Demand
17
18	\| Input \| SKILL \| Template \| References \|
19	\| --------------- \| ----- \| -------- \| ---------------------- \|
20	\| Quick Q (<50 w) \| full \| ✓ \| if error/level unclear \|
21	\| Standard Q \| full \| ✓ \| if error/level unclear \|
22	\| Batch (≥5) \| full \| ✓ \| both required \|
23	\| Weak spot \| full \| ✓ \| diagnostic-rubric \|
24
25	## Decision Rules (priority: batch > type)
26
27	\| Condition \| Action \|
28	\| ---------- \| --------------------------------------- \|
29	\| Single Q \| 5 steps \|
30	\| ≥5 Qs \| + batch statistics \|
31	\| weak_spot \| Strengthen error analysis \|
32	\| Image/OCR \| Verify OCR first \|
33
34
35	## Self-Check (1-2 fails → fix inline · 3+ → regenerate)
36
37	\| Check \| Rule \|
38	\| -------- \| ----------------------------------------------------------------------------------- \|
39	\| Filled \| All fields or "N/A" \|
40	\| Tiers \| `[*]/[]/[*]` (no ★) \|
41	\| Time \| 1-2 min recall, 3-5 min apply \|
42	\| Depth \| Surface verb ≠ Hidden verb (see[blooms-taxonomy.md](./references/blooms-taxonomy.md)) \|
43	\| Order \| High-frequency first, not textbook \|
44	\| Coverage \| Template ≥80% of 5 variants (4/5 pass) \|
45	\| Errors \| `symptom → cause → prevention`, concrete checklist \|
46	\| Truth \| No fabrication; mark "approx." \|
47
48	## Top 3 Errors (3D: A frequency + B severity + C preventability, 1-3 each)
49
50	\| Dim \| 1 \| 2 \| 3 \|
51	\| ----------- \| ------- \| ---------- \| ------------------ \|
52	\| A \| Rare \| Occasional \| Dominant \|
53	\| B \| 1-2 pts \| 3-5 pts \| Full Q / cascade \|
54	\| C \| Insight \| Partial \| Fully templateable \|
55
56	Tie-break: C > A. Pick top 3. Output: `symptom → cause → prevention` (concrete checklist).
57
58	## Positive Directives
59
60	- Order by exam frequency, not textbook
61	- `[*] / [] / [*]` consistent · `→` for cause-effect
62	- Hidden intent deeper than surface
63	- Template ≥80% · Errors actionable · Time budget mandatory
64	- Bullets/tables > prose · Supplement study, not replace · No fabrication
65
66	## Limitations (do NOT use)
67
68	- Open-ended essays / writing prompts
69	- Live exam proctoring
70	- Subjects outside STEM + basic humanities
71	- Source images with unclear OCR
72	- Single factual lookup · User wants the actual answer
73
74	Caveat: Tiers reflect common patterns, not guaranteed outcomes. Cross-check with official syllabus.
75
76	## File Map
77
78	\| File \| Purpose \| Load \|
79	\| ------------------------------------------------------ \| -------------------- \| ---------------------- \|
80	\| [SKILL.md](./SKILL.md) \| Pipeline + rules \| Always \|
81	\| [output-template.md](./assets/output-template.md) \| Output structure \| When generating \|
82	\| [blooms-taxonomy.md](./references/blooms-taxonomy.md) \| L1-L6 levels \| When classifying level \|
83	\| [diagnostic-rubric.md](./references/diagnostic-rubric.md) \| 5 error categories \| When building errors \|
84	\| [quantified-rubric.md](./references/quantified-rubric.md) \| Quantified standards \| When self-checking \|

People who liked this also viewed

6 picks

Type