Long Form Writer

Generate comprehensive long-form articles and reports with structured depth.

When to Use This Skill

• Create tutorials exceeding 2000 words
• Write research reports with multiple sections
• Generate analysis documents with case studies
• Transform raw data into structured narratives
• Expand outline into full content

Core Workflow (4 Layers)

L1: Structure (大纲规划)

Define the article architecture:

## Article Blueprint

**Type**: [tutorial | report | analysis | guide]
**Target Length**: [3000 | 5000 | 8000 | 10000] words
**Audience**: [technical | general | executive]

### Chapter Plan
1. [Title] - [Word count] - [Key points]
2. [Title] - [Word count] - [Key points]
...

L2: Expand (章节展开)

Generate each chapter independently:

For each chapter:
1. Write 500-1500 words
2. Include: concept + example + application
3. Add subsections if needed (recursive)
4. Mark [EXPAND] for sections needing more depth

L3: Enrich (素材增强)

Add supporting materials:

• Cases: Real user examples from chat history
• Data: Statistics, comparisons, trends
• Visuals: Screenshot suggestions, diagram descriptions
• FAQ: Common questions and answers

L4: Polish (润色检查)

Apply quality standards:

• Remove AI filler phrases
• Check for 3-part structures → convert to 2 or 4
• Vary sentence length
• Add personal insights
• Verify logical flow between chapters

Usage Patterns

Pattern 1: From Data to Report

# Input: analysis.json
# Output: comprehensive report

long-form-writer generate \
  --input data.json \
  --type report \
  --template structured \
  --output report.md

Pattern 2: From Outline to Article

# Input: outline.md
# Output: full tutorial

long-form-writer expand \
  --outline outline.md \
  --depth 3 \
  --word-count 5000 \
  --output article.md

Pattern 3: Chat History to Guide

# Extract cases from chat, generate guide

long-form-writer guide \
  --chat-history chat.json \
  --topic "OpenClaw Usage" \
  --output guide.md

Integration with Other Skills

Skill	Usage
feishu-chat-extractor	Extract real cases from group chats
summarize	Compress reference materials
nano-pdf	Export final PDF

Output Formats

• Markdown (default): Editable, version control friendly
• PDF: Professional delivery
• Feishu Doc: Collaborative editing

Quality Checklist

Before delivery, verify:

1. Each chapter has concrete examples
2. Total word count meets target
3. No section is purely theoretical
4. Data/sources are cited
5. Conclusion ties back to introduction
6. Formatting is consistent

Examples

See references/examples.md for sample outputs.