Learning Content Creator

Transform multi-model research materials into structured learning content, with bilingual support (English + Korean).

Quick Start

Workflow Overview

Research Materials → Learning Path (EN) → Translation (KO) → Frontmatter

Workflow Routing

Note: Phase 4 (Frontmatter) uses the doc-frontmatter schema. The calling command coordinates the multi-skill pipeline.

Phase 1: ANALYZE Research

Goal: Understand and synthesize multi-model research.

Input Structure

docs/NN-topic/
├── research/
│   ├── 00-research-prompt.en.md   # Original prompt
│   ├── 01-claude.en.md            # Claude's response
│   ├── 02-gpt/                    # GPT's response (may be multi-file)
│   │   ├── 01-concepts.en.md
│   │   ├── 02-relationships.en.md
│   │   └── ...
│   └── 03-gemini.en.md            # Gemini's response

Analysis Checklist

Output: Content Outline

markdown
1# Learning Content Outline
2
3## Modules (6-8 recommended)
4
5| # | Module | Topics | Sources |
6|---|--------|--------|---------|
7| 1 | Fundamentals | Definitions, core concepts | All models |
8| 2 | Relationships | How parts connect | GPT, Claude |
9| ... | ... | ... | ... |
10
11## Key Insights by Source
12
13- **Claude**: [unique insights]
14- **GPT**: [unique insights]
15- **Gemini**: [unique insights]
16
17## Synthesis Strategy
18
19[How to combine insights without redundancy]

Phase 2: CREATE English Content

Goal: Create structured learning modules in English.

Directory Structure

docs/NN-topic/
└── learning/
    ├── README.en.md           # Course overview, learning path
    ├── 01-module-name.en.md   # Module 1
    ├── 02-module-name.en.md   # Module 2
    └── ...

Module Template

markdown
1# Module N: Title
2
3> One-sentence module summary
4
5## Learning Objectives
6
7After completing this module, you will:
8- [Objective 1]
9- [Objective 2]
10- [Objective 3]
11
12---
13
14## N.1 First Section
15
16[Content]
17
18## N.2 Second Section
19
20[Content]
21
22---
23
24## Key Takeaways
25
26- [Takeaway 1]
27- [Takeaway 2]
28- [Takeaway 3]
29
30## Exercises
31
32### Exercise N.1: [Name]
33
34[Exercise description]
35
36---
37
38## Next Steps
39
40Continue to [Module N+1: Title](./0N+1-title.en.md)

README Template

markdown
1# Course Title
2
3> Course tagline (one sentence)
4
5## Course Overview
6
7[2-3 paragraph description]
8
9## Who This Is For
10
11- [Audience 1]
12- [Audience 2]
13
14## Prerequisites
15
16- [Prerequisite 1]
17- [Prerequisite 2]
18
19---
20
21## Course Modules
22
23| # | Module | Duration | Description |
24|---|--------|----------|-------------|
25| 1 | [Title](./01-name.en.md) | NN min | Description |
26| 2 | [Title](./02-name.en.md) | NN min | Description |
27
28**Total Time:** ~N hours
29
30---
31
32## Learning Path
33
34### Beginner Track
35[Visual flow diagram]
36
37### Advanced Track
38[Visual flow diagram]
39
40---
41
42## Source Materials
43
44| Source | Description |
45|--------|-------------|
46| [Research Prompt](../research/00-research-prompt.en.md) | Original prompt |
47| [Claude](../research/01-claude.en.md) | Claude's response |
48| [GPT](../research/02-gpt/) | GPT's response |
49| [Gemini](../research/03-gemini.en.md) | Gemini's response |

Writing Guidelines

Phase 3: TRANSLATE to Korean

Goal: Create high-quality Korean translations.

Naming Convention

Translation Guidelines

Do NOT Translate

• Code snippets
• File paths
• Command examples
• Technical identifiers (kebab-case names, etc.)

Translation Checklist

• All .en.md files have .ko.md counterparts
• Internal links updated to .ko.md versions
• Technical terms consistently handled
• Tables and diagrams preserved
• Exercises and examples localized where appropriate

Phase 4: ADD Frontmatter

Goal: Add YAML frontmatter to all learning documents.

Frontmatter Schema

Use the frontmatter schema from doc-frontmatter skill (see .claude/skills/doc-frontmatter/references/schema.md).

Frontmatter Template for Learning Content

yaml
1---
2title: "Module Title"
3description: "50-160 char summary of what this module teaches"
4type: tutorial
5tags: [AI, Architecture, BestPractice]
6order: 1
7depends_on: [./prerequisite-module.en.md]
8related: [./related-module.en.md]
9---

Type Selection for Learning

Execution

For each file:

1. Extract title from H1
2. Generate description from first paragraph
3. Determine type based on content
4. Select relevant tags (max 5)
5. Set order from filename prefix
6. Add depends_on/related if applicable

Quality Checklist

Phase 1: Analyze

• All research files read
• Key insights extracted from each model
• Conflicts identified and resolved
• Content outline created

Phase 2: Create EN

• All modules follow template
• Learning objectives are measurable
• Content synthesizes (not copies) sources
• Exercises included in each module
• Links work correctly

Phase 3: Translate KO

• All files translated
• Technical terms consistent
• Links updated to .ko.md
• Natural Korean (not machine-translation quality)

Phase 4: Frontmatter

• All files have frontmatter
• Required fields present (title, description, type)
• Tags from controlled vocabulary
• Dependencies correctly specified

Example Output

See docs/01-structure-organizer/learning/ for a complete example:

docs/01-structure-organizer/learning/
├── README.en.md              # Course index (EN)
├── README.ko.md              # Course index (KO)
├── 01-fundamentals.en.md     # Module 1 (EN)
├── 01-fundamentals.ko.md     # Module 1 (KO)
├── 02-relationships.en.md    # Module 2 (EN)
├── 02-relationships.ko.md    # Module 2 (KO)
├── 03-decision-framework.en.md
├── 03-decision-framework.ko.md
├── 04-templates.en.md
├── 04-templates.ko.md
├── 05-examples.en.md
├── 05-examples.ko.md
├── 06-anti-patterns.en.md
└── 06-anti-patterns.ko.md