pptx

PPT Presentation Processing Expert Real Skill

One-sentence description

Automatically generate beautiful PPTs from HTML, supporting presentation editing, comment management, and content extraction.

Presentation Creation & Editing - Creating presentations from HTML, editing content, managing comments, extracting text.

Use Cases

Applicable：

Automatically generate standardized PPT reports from data/HTML
Create beautiful presentations (sales, training, reporting)
Extract PPT text, comments, speaker notes
Modify slide content, layout, theme
Batch generate multiple presentations

Not Applicable：

Complex animation production (use professional software)
Real-time collaborative editing (use Google Slides for real-time collaboration)

Core Workflow

用户需求 → 内容分析 → 配色选择 → 工具选择 → 生成 PPT → 验证输出
User request → Content analysis → Color palette → Tool selection → Generate PPT → Validate

Task Classification：

Creating Presentations

HTML to PPT Workflow (RECOMMENDED)
- Write HTML slides with styling
- Convert using html2pptx library
- Accurate positioning and formatting
- Web-safe fonts only (Arial, Helvetica, Georgia, etc.)
Python-PPTX Workflow
- Programmatic slide creation
- Batch generation from data
- Template-based automation

Reading & Analysis

Text extraction → markitdown (convert to markdown)
Comments & Notes → Unpack OOXML and read XML
Design analysis → Extract theme, colors, fonts from XML

Editing Existing

Simple edits → python-pptx library
Complex edits → Unpack OOXML, modify XML, repack

Task Completion Criteria

Must Satisfy(all are required):

PPT opens and plays normally
Slide content is clear and readable
Coordinated color scheme with clear visual hierarchy
Reasonable layout with no crowded text

Quality Rating：

⭐⭐⭐⭐⭐ Excellent - Exquisite design + complete content + automated generation
⭐⭐⭐ Pass - Correct content + usable PPT + standard formatting
⭐ Fail - PPT is corrupted or content is incorrect

Reference Materials (for AI use)

Type	Path	Description
Core Documentation	`docs/00-SKILL-完整制作指南.md`	Complete PPT creation guide (~500 lines)
OOXML Library	`ooxml/`	Low-level XML operations
Tool Scripts	`scripts/unpack.py`	Unpack PPTX files
Tool Scripts	`scripts/pack.py`	Pack PPTX files

Critical Principles (AI Must Read)

1. Design First - Content-Informed Approach

CRITICAL: Before creating any presentation, analyze the content:

Consider subject matter: What is this about? What tone/industry/mood?
Check branding: If company mentioned, consider brand colors
Match palette to content: Select colors reflecting the subject
State your approach: Explain design choices BEFORE writing code

2. Web-Safe Fonts Only

Allowed fonts: Arial, Helvetica, Times New Roman, Georgia, Courier New, Verdana, Tahoma, Trebuchet MS, Impact

No custom fonts: They won't render correctly
Font pairing: Use max 2-3 fonts per presentation

3. Visual Hierarchy

Extreme size contrast: 72pt headlines vs 11pt body text
Weight variation: Bold headers, normal body, light captions
Color emphasis: Use accent color for key information
Clear structure: Title → Main content → Supporting details

4. Chart & Table Layout Rules

PREFERRED - Two-column layout:

Header spanning full width
Text/bullets in one column (40%) + Chart/table in other (60%)
Better balance and readability

Alternative - Full-slide layout:

Chart/table takes entire slide for maximum impact

NEVER vertically stack:

❌ Do NOT place charts/tables below text in single column
❌ Causes poor readability and layout issues

5. Color Palette Selection (18 color schemes)

Think beyond defaults: What colors genuinely match this topic? Consider multiple angles: Topic, industry, mood, energy, audience, brand Be adventurous: Healthcare doesn't have to be green, finance doesn't have to be navy

18 Selected Color Palettes

Business & Professional

Classic Blue: Navy (#1C2833), Slate (#2E4053), Silver (#AAB7B8), Off-white (#F4F6F6)
Black & Gold: Gold (#BF9A4A), Black (#000000), Cream (#F4F6F6)
Charcoal & Red: Charcoal (#292929), Red (#E33737), Light Gray (#CCCBCB)

Creative & Vibrant

Teal & Coral: Teal (#5EA8A7), Deep Teal (#277884), Coral (#FE4447), White (#FFFFFF)
Retro Rainbow: Purple (#722880), Pink (#D72D51), Orange (#EB5C18), Amber (#F08800), Gold (#DEB600)
Vibrant Orange: Orange (#F96D00), Light Gray (#F2F2F2), Charcoal (#222831)
Bold Red: Red (#C0392B), Bright Red (#E74C3C), Orange (#F39C12), Yellow (#F1C40F), Green (#2ECC71)

Natural & Warm

Sage & Terracotta: Sage (#87A96B), Terracotta (#E07A5F), Cream (#F4F1DE), Charcoal (#2C2C2C)
Forest Green: Black (#191A19), Green (#4E9F3D), Dark Green (#1E5128), White (#FFFFFF)
Vintage Earthy: Mustard (#E3B448), Sage (#CBD18F), Forest Green (#3A6B35), Cream (#F4F1DE)
Cream & Forest: Cream (#FFE1C7), Forest Green (#40695B), White (#FCFCFC)

Elegant & Soft

Warm Blush: Mauve (#A49393), Blush (#EED6D3), Rose (#E8B4B8), Cream (#FAF7F2)
Coastal Rose: Old Rose (#AD7670), Beaver (#B49886), Eggshell (#F3ECDC), Ash Gray (#BFD5BE)
Pink & Purple: Pink (#F8275B), Coral (#FF574A), Rose (#FF737D), Purple (#3D2F68)

Luxury & Bold

Burgundy Luxury: Burgundy (#5D1D2E), Crimson (#951233), Rust (#C15937), Gold (#997929)
Deep Purple & Emerald: Purple (#B165FB), Dark Blue (#181B24), Emerald (#40695B), White (#FFFFFF)

Fresh & Modern

Lime & Plum: Lime (#C5DE82), Plum (#7C3A5F), Coral (#FD8C6E), Blue-Gray (#98ACB5)
Orange & Turquoise: Light Orange (#FC993E), Grayish Turquoise (#667C6F), White (#FCFCFC)

Quick Command Reference

Text Extraction

# Convert PPTX to markdown
python -m markitdown presentation.pptx

Create Simple Presentation

from pptx import Presentation

prs = Presentation()
slide = prs.slides.add_slide(prs.slide_layouts[0])  # Title slide
title = slide.shapes.title
subtitle = slide.placeholders[1]

title.text = "Hello World!"
subtitle.text = "My First Presentation"

prs.save("output.pptx")

Unpack PPTX for Analysis

# Unpack PPTX to access raw XML
python ooxml/scripts/unpack.py presentation.pptx output_dir

# Key files to examine:
# ppt/presentation.xml - Main metadata
# ppt/slides/slide1.xml - Individual slide contents
# ppt/theme/theme1.xml - Theme colors and fonts
# ppt/comments/ - Comments
# ppt/notesSlides/ - Speaker notes

Repack After Editing

# Pack modified XML back to PPTX
python ooxml/scripts/pack.py output_dir new_presentation.pptx

HTML to PPT Workflow

# Step 1: Write HTML slides
html_content = """
<!DOCTYPE html>
<html>
<head>
    <style>
        .slide { width: 960px; height: 720px; padding: 40px; }
        h1 { font-family: Arial; font-size: 48px; color: #1C2833; }
        p { font-family: Arial; font-size: 24px; color: #2E4053; }
    </style>
</head>
<body>
    <div class="slide">
        <h1>Welcome</h1>
        <p>This is my presentation</p>
    </div>
</body>
</html>
"""

# Step 2: Convert to PPTX
from html2pptx import html2pptx
with open("slides.html", "w") as f:
    f.write(html_content)
html2pptx("slides.html", "output.pptx")

Dependencies Installation

Python Dependencies

pip install -r requirements.txt
# Includes: python-pptx, html2pptx, markitdown

Common Scenarios

Scenario 1: Create Professional Title Slide

from pptx import Presentation
from pptx.util import Inches, Pt
from pptx.dml.color import RGBColor

prs = Presentation()
slide = prs.slides.add_slide(prs.slide_layouts[6])  # Blank layout

# Add title
left = Inches(1)
top = Inches(2.5)
width = Inches(8)
height = Inches(1.5)
title_box = slide.shapes.add_textbox(left, top, width, height)
title_frame = title_box.text_frame
title_frame.text = "2024 Sales Report"

# Format title
title_para = title_frame.paragraphs[0]
title_para.font.size = Pt(72)
title_para.font.bold = True
title_para.font.color.rgb = RGBColor(28, 40, 51)  # Navy

# Add subtitle
subtitle_box = slide.shapes.add_textbox(left, Inches(4), width, Inches(0.5))
subtitle_frame = subtitle_box.text_frame
subtitle_frame.text = "Q4 Performance Review"
subtitle_para = subtitle_frame.paragraphs[0]
subtitle_para.font.size = Pt(24)
subtitle_para.font.color.rgb = RGBColor(46, 64, 83)  # Slate

prs.save("title_slide.pptx")

Scenario 2: Extract All Speaker Notes

# Step 1: Unpack PPTX
python ooxml/scripts/unpack.py presentation.pptx unpacked/

# Step 2: Extract notes using grep
cd unpacked/ppt/notesSlides/
grep -r "a:t>" . | sed 's/<[^>]*>//g' > ../../all_notes.txt

Scenario 3: Batch Generate Slides from Data

from pptx import Presentation
from pptx.util import Inches, Pt

data = [
    {"name": "Product A", "sales": 1200, "growth": "15%"},
    {"name": "Product B", "sales": 980, "growth": "8%"},
    {"name": "Product C", "sales": 1450, "growth": "22%"}
]

prs = Presentation()

for item in data:
    slide = prs.slides.add_slide(prs.slide_layouts[1])  # Title and content
    title = slide.shapes.title
    title.text = item["name"]
    
    body = slide.placeholders[1]
    tf = body.text_frame
    tf.text = f"Sales: ${item['sales']}M"
    
    p = tf.add_paragraph()
    p.text = f"Growth: {item['growth']}"
    p.level = 1

prs.save("batch_report.pptx")