nachafrommars/atydam-epub-v3.0/atydam-epub-v3-0

Claude Code and agent skill by nachafrommars/atydam-epub-v3.0

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Automated PDF→EPUB conversion pipeline v3.0 with GPT-STYLE bold emphasis. Same 9-phase pipeline as v2.0 PLUS Phase B10 (AI bold emphasis). Produces scannable, easy-to-read EPUB with ~1,500 bold keywords per 400-page book, following ChatGPT writing style. For plain text version without bold, use atyDam-epub v2.0.

Install

npx vskill@latest install nachafrommars/atydam-epub-v3.0/atydam-epub-v3-0

SKILL.md

atyDam-epub v3.0 — PDF→EPUB with GPT-STYLE Bold Emphasis

When to use this skill

When user wants to convert a PDF book to easy-read EPUB with bold keywords (GPT-style):

  • PDF with broken font encoding (Vietnamese + Pali Buddhist texts)
  • Books with complex tables, figures, footnotes
  • Want bold emphasis for scanability (ChatGPT style)
  • Want keywords highlighted for quick reference

For plain text version without bold, use atyDam-epub v2.0 instead.

What is GPT-Style?

GPT-Style = AI reads & understands content, then bolds important keywords per ChatGPT writing conventions:

  1. Definitions — key concepts being explained
  2. Lists — important items in series
  3. Numbers — quantities emphasized (12 tâm, 18 tâm)
  4. Conclusions — summary sentences
  5. Proper names — Phật học terms & entities
  6. Contrasts — comparative phrases ("different from...", "unlike...")
  7. Key instructions — "remember..." or "note that..."
  8. Scan keywords — quick-reference terms

Result: ~1,500 bold patterns per 400-page book → 1-3 bold per paragraph → dễ scan, dễ đọc.

Pipeline Overview — 10 Phases + Audit Loop + GPT-Style

PDF input
  │
  ├─► Phase A: EXTRACT (PyMuPDF)
  │     • Extract images, render cover, chapters/TOC
  │     • Output: extract_meta.json
  │
  ├─► Phase 1: TEXT RECOVERY (Gemini Flash)
  │     • Fix Vietnamese diacritics + Pali font damage
  │     • thinkingBudget=0 (CRITICAL)
  │
  ├─► Phase B2-B3: STRUCTURE PREP (regex)
  │     • Paragraph reconstruct + inline list split
  │
  ├─► Phase B6-B7: AI STRUCTURE (Gemini)
  │     • Heading hierarchy + semantic paragraphs
  │
  ├─► Phase B9: TABLE VISION (Gemini Vision)
  │     • PDF image → HTML table with rowspan/colspan
  │
  ├─► Phase B5: PROOFREAD (Gemini)
  │     • Spelling + Phật học terminology
  │
  ├─► Phase B10: GPT-STYLE BOLD (Gemini)
  │     ◄─── NEW IN v3.0 ────────────
  │     • AI reads & understands paragraphs
  │     • Bold keywords per 8 ChatGPT rules
  │     • ~1,500 patterns per 400 pages
  │     • Cost: ~$0.38
  │
  ├─► Phase C-D: MARKUP + BUILD (Python + Pandoc)
  │     • YAML front matter + Pali italic
  │     • Cover + CSS + EPUB 3.2
  │
  └─► AUDIT LOOP (7 layers)
        • Structural, semantic, spelling, tables, images, cross-ref, meta

Step-by-step Execution

Step 1: Setup workspace
bash
WORKDIR="/root/.openclaw/workspace/<book-name>-epub"
mkdir -p "$WORKDIR/{chunks,recovered,images}"
cd "$WORKDIR"
cp scripts/* .  # Copy all phase scripts including phase_b10_bold_emphasis.py
Step 2: Run pipeline
bash
export GEMINI_API_KEY="..."
 
# Phase A: Extract
python3 phase_a_extract.py
 
# Phase 1: Recovery
python3 pipeline_parallel.py
 
# Phase B2-B3: Structure prep
python3 phase_b2_paragraphs.py
python3 phase_b3_smart_list.py
 
# Phase B6-B7: AI structure  
python3 phase_b6_smart_structure.py
python3 phase_b7_perfect_structure.py
 
# Phase B9: Vision tables
python3 phase_b9_tables_vision.py
 
# Phase B5: Proofread
python3 phase_b5_proofread.py
 
# ─── Phase B10: GPT-STYLE BOLD (NEW) ───
python3 phase_b10_bold_emphasis.py
 
# Phase C-D: Build
python3 phase_c_markup.py
python3 phase_d_build.py
 
# Audit loop
python3 audit_loop.py
python3 audit_v2.py
 
# Final build
python3 phase_d_build.py
Step 3: Validate
bash
java -jar /usr/share/java/epubcheck.jar TrietHoc-*.epub
# Expected: 0 fatals / 0 errors / 0 warnings

Cost per book (~420 pages)

PhaseCost
Phase 1 Recovery$0.54
Phase 2 Tables$0.20
Phase 5 Proofread$0.40
Phase 7 Structure$0.42
Phase 8 Perfect$0.42
Phase B10 Bold$0.38 ← v3.0 only
Audit V1$0.79
Audit V2$0.38
TOTAL v3.0$3.53 (~85,000 VNĐ)

vs v2.0: $3.15 (no B10 phase)

Quality Indicators (v3.0)

  • ✅ Avg paragraph length 150-380 chars
  • ✅ 9 chapters proper hierarchy
  • ✅ All tables HTML with thead/tbody/th/td
  • ✅ Italic Pali wrapping (~1,100+ terms)
  • ✅ epubcheck: 0 errors / 0 warnings
  • ✅ Images embedded (~30 figures)
  • ✅ Cover from PDF page 1
  • ✅ ~1,500 bold patterns (GPT-style)
  • ✅ 1-3 bold per paragraph (readable, not overwhelming)
  • ✅ No bold in tables, headings, Pali italic

Difference: v2.0 vs v3.0

Featurev2.0v3.0
Text recovery✅✅
Structure detection✅✅
Tables rebuild✅✅
Proofreading✅✅
GPT-Style bold❌ NO✅ YES
Bold patterns0~1,500
Cost per book~$3.15~$3.53

Files Structure

<book>-epub/
├── chunks/              
├── recovered/           
├── images/              
├── extract_meta.json    
├── cover.png            
├── style-v2-premium.css 
├── pipeline_parallel.py 
├── phase_a_extract.py   
├── phase_b2_paragraphs.py  
├── phase_b3_smart_list.py  
├── phase_b5_proofread.py   
├── phase_b6_smart_structure.py  
├── phase_b7_perfect_structure.py 
├── phase_b9_tables_vision.py    
├── phase_b10_bold_emphasis.py  ◄─── v3.0 NEW
├── phase_c_markup.py    
├── phase_d_build.py     
├── audit_loop.py        
├── audit_v2.py          
├── audit-log.json       
└── TrietHoc-A-Ty-Dam-VN.epub  # Final output WITH BOLD

Trigger Auto-run

When user explicitly requests v3.0 or GPT-style:

  • "convert PDF sang EPUB với gpt-style"
  • "pdf to epub gpt-style"
  • "epub với bold"
  • "chatgpt style epub"
  • etc.

Em will run v3.0 (includes Phase B10).

For standard conversion without bold, default is v2.0.

Memory Note

atyDam-epub v3.0 is the GPT-style bold version.

  • Produced from work done 16/05/2026
  • Phase B10 (AI bold emphasis) added to standard v2.0 pipeline
  • V11 of Triết Học A-Tỳ-Đàm: example v3.0 output (1,524 bold patterns)

For plain text version, see atyDam-epub v2.0 skill.

Security

  • Security scanPassed, 95 of 100Pattern checks on the skill's files.
  • Model reviewPassed

See the full security report

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

See evaluation results

Source

  • SKILL.md in the repositoryCheckingSKILL.md

Popularity

7-day trend score
5

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/nachafrommars/atydam-epub-v3.0/atydam-epub-v3-0/badge)](https://verified-skill.com/skills/nachafrommars/atydam-epub-v3.0/atydam-epub-v3-0)

Something wrong with this skill or its page? Tell us.