Claude-skill-registry draftify-ppt
This skill generates planning documents (기획서) in PowerPoint format from analyzed screen data and project artifacts. It should be used by the auto-draft-orchestrator agent during Phase 4 to create the final PPT output. Not intended for direct user invocation.
git clone https://github.com/majiayu000/claude-skill-registry
T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/draftify-ppt" ~/.claude/skills/majiayu000-claude-skill-registry-draftify-ppt && rm -rf "$T"
skills/data/draftify-ppt/SKILL.mdDraftify PPT Generator
Overview
This skill transforms analyzed project data into a structured PowerPoint planning document (기획서) following the standard 10-section format. It is invoked by the auto-draft-orchestrator during Phase 4 of the document generation workflow.
Input Requirements
The skill expects the following files to exist in the project's output directory:
outputs/<project-name>/ ├─ screenshots/ # Captured screen images ├─ analysis/ │ └─ analyzed-structure.json # Consolidated analysis data ├─ sections/ │ ├─ 05-glossary.md # 용어 정의 │ ├─ 06-policy-definition.md # 정책 정의 │ ├─ 07-process-flow.md # 프로세스 흐름 │ └─ 08-screen-definition.md # 화면 정의 └─ validation/ └─ validation-report.md # Quality validation results
Output
Generates
final-draft.pptx in the project output directory:
outputs/<project-name>/ └─ final-draft.pptx
Document Structure
The generated PPT follows the 10-section structure defined in
references/auto-draft-guideline.md:
| Section | Content |
|---|---|
| 1. 표지 | Cover with project metadata |
| 2. 변경 이력 | Revision history table |
| 3. 목차 | Table of contents with screen IDs |
| 4. 섹션 타이틀 | Section divider pages |
| 5. 용어 정의 | Glossary terms from |
| 6. 정책 정의 | Policies with POL-* IDs from |
| 7. 프로세스 흐름 | Process flow from |
| 8. 화면 정의 | Screen definitions with SCR-* IDs from |
| 9. 참고 문헌 | Reference documents |
| 10. EOD | End of document marker |
Generation Workflow
- Read analyzed data: Load
and section markdown filesanalyzed-structure.json - Parse screen definitions: Extract screen metadata, screenshots, and element definitions
- Generate cover slide: Use project name, version, and current date
- Generate TOC: Create clickable table of contents
- Generate section slides: Process each section markdown into slides
- Insert screenshots: Embed captured screenshots in screen definition slides
- Apply template styling: Use
as baseassets/ppt_template.pptx - Save output: Write
final-draft.pptx
ID Scheme Compliance
All IDs must follow the scheme defined in the guideline:
- Policy IDs:
(e.g., POL-AUTH-001)POL-{CATEGORY}-{SEQ} - Screen IDs:
(e.g., SCR-001)SCR-{SEQ} - Element IDs:
(e.g., BTN-001, FORM-001){TYPE}-{SEQ} - API IDs:
(e.g., API-001)API-{SEQ}
Screen Definition Slide Layout
Each screen definition uses 1-2 slides with this structure:
Slide 1 (Required):
- Screen ID and name (header)
- Screenshot image (left 60%)
- Basic info panel (right 40%): purpose, entry/exit conditions
Slide 2 (If needed):
- UI element table
- Process flow within screen
- Related policies (POL-* references)
Usage by Orchestrator
The auto-draft-orchestrator invokes this skill via Task tool:
Task: Generate final PPT document Input: outputs/<project-name>/ directory path Timeout: 10 minutes
Error Handling
- Missing section files: Generate placeholder slide with warning
- Missing screenshots: Use placeholder image with screen ID
- Invalid IDs: Log warning, continue generation
- Template errors: Fall back to basic slide layout
Resources
scripts/
: Main PPT generation script using python-pptxgenerate_ppt.py
references/
: Complete specification for document structure and ID schemesauto-draft-guideline.md
assets/
: PowerPoint template with predefined layouts and stylingppt_template.pptx
: Bold font for headersJOURNEYITSELF-BOLD 3.TTF
: Regular font for body textJOURNEYITSELF-REGULAR 3.TTF
: Light font for captionsJOURNEYITSELF-LIGHT 3.TTF
Dependencies
- Python 3.8+
- python-pptx library
- Pillow (for image processing)