Powerpoint

PowerPoint Skill

Generates, updates, and manages PowerPoint slide decks using python-pptx with YAML-driven content and styling definitions.

Overview

This skill provides Python scripts that consume YAML configuration files to produce PowerPoint slide decks. Each slide is defined by a content.yaml file describing its layout, text, and shapes. A style.yaml file defines dimensions, template configuration, layout mappings, metadata, and defaults.

SKILL.md covers technical reference: prerequisites, commands, script architecture, API constraints, and troubleshooting. For conventions and design rules (element positioning, visual quality, color and contrast, contextual styling), follow pptx.instructions.md.

Prerequisites

PowerShell

The Invoke-PptxPipeline.ps1 script handles virtual environment creation and dependency installation automatically via uv sync. Requires uv, Python 3.11+, and PowerShell 7+.

Installing uv

If uv is not installed:

# macOS / Linux
curl -LsSf https://astral.sh/uv/install.sh | sh

# Windows
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

# Via pip (fallback)
pip install uv

System Dependencies (Export and Validation)

The Export and Validate actions require LibreOffice for PPTX-to-PDF conversion and optionally pdftoppm from poppler for PDF-to-JPG rendering. When pdftoppm is not available, PyMuPDF handles the image rendering.

The Validate action's vision-based checks require the GitHub Copilot CLI for model access.

# macOS
brew install --cask libreoffice
brew install poppler        # optional, provides pdftoppm

# Linux
sudo apt-get install libreoffice poppler-utils

# Windows (winget preferred, choco fallback)
winget install TheDocumentFoundation.LibreOffice
# choco install libreoffice-still      # alternative
# poppler: no winget package; use choco install poppler (optional, provides pdftoppm)

Copilot CLI (Vision Validation)

The validate_slides.py script uses the GitHub Copilot SDK to send slide images to vision-capable models. The Copilot CLI must be installed and authenticated:

# Install Copilot CLI
npm install -g @github/copilot-cli

# Authenticate (uses the same GitHub account as VS Code Copilot)
copilot auth login

# Verify
copilot --version

Required Files

• style.yaml — Dimensions, defaults, template configuration, and metadata
• content.yaml — Per-slide content definition (text, shapes, images, layout)
• (Optional) content-extra.py — Custom Python for complex slide drawings

Content Directory Structure

All slide content lives under the working directory's content/ folder:

content/
├── global/
│   ├── style.yaml              # Dimensions, defaults, template config, and theme metadata
│   └── voice-guide.md          # Voice and tone guidelines
├── slide-001/
│   ├── content.yaml            # Slide 1 content and layout
│   └── images/                 # Slide-specific images
│       ├── background.png
│       └── background.yaml     # Image metadata sidecar
├── slide-002/
│   ├── content.yaml            # Slide 2 content and layout
│   ├── content-extra.py        # Custom Python for complex drawings
│   └── images/
│       └── screenshot.png
├── slide-003/
│   ├── content.yaml
│   └── images/
│       ├── diagram.png
│       └── diagram.yaml
└── ...

Global Style Definition (style.yaml)

The global style.yaml defines dimensions, template configuration, layout mappings, metadata, and defaults. Color and font choices are specified per-element in each slide's content.yaml rather than centralized in the style file.

See the [style.yaml template](style-yaml-template.md) for the full template, field reference, and usage instructions.

Per-Slide Content Definition (content.yaml)

Each slide's content.yaml defines layout, text, shapes, and positioning. All position and size values are in inches. Color values use #RRGGBB hex format or @theme_name references.

Text contract: markdown-like list lines in textbox.text and shape.text are interpreted as PowerPoint lists during rendering. Unordered markers (-, +, *) become bulleted paragraphs, ordered markers (1., 1)) become auto-numbered paragraphs, and leading indentation maps to paragraph level.

See the [content.yaml template](content-yaml-template.md) for the full template, supported element types, supported shape types, and usage instructions.

Complex Drawings (content-extra.py)

When a slide requires complex drawings that cannot be expressed through content.yaml element definitions, create a content-extra.py file in the slide folder. The render() function signature is fixed. The build script calls it after placing standard content.yaml elements.

See the [content-extra.py template](content-extra-py-template.md) for the full template, function parameters, and usage guidelines.

Security Validation

Before executing a content-extra.py file, the build script performs AST-based static analysis to reject dangerous code. Validation runs automatically unless the --allow-scripts flag is passed.

Allowed imports:

• pptx and all pptx.* submodules
• Safe standard-library modules (e.g., math, copy, json, re, pathlib, collections, itertools, functools, typing, enum, dataclasses, decimal, fractions, string, textwrap)

Blocked imports:

• subprocess, os, shutil, socket, ctypes, signal, multiprocessing, threading, http, urllib, ftplib, smtplib, imaplib, poplib, xmlrpc, webbrowser, code, codeop, compileall, py_compile, zipimport, pkgutil, runpy, ensurepip, venv, sqlite3, tempfile, shelve, dbm, pickle, marshal, importlib, sys, telnetlib
• Any third-party package not on the allowlist

Blocked builtins:

• Dangerous: eval, exec, __import__, compile, breakpoint
• Indirect bypass: getattr, setattr, delattr, globals, locals, vars

Runtime namespace restriction:

Even after AST validation passes, the executed module runs in a restricted namespace where __builtins__ is limited to safe builtins only. The dangerous and indirect-bypass builtins listed above are removed from the module namespace before execution (__import__ is kept because the import machinery requires it; the AST checker blocks direct __import__() calls).

--allow-scripts flag:

Pass --allow-scripts to skip AST validation and namespace restriction for trusted content. This flag is required when a content-extra.py script legitimately needs blocked imports or builtins.

python scripts/build_deck.py \
  --content-dir content/ \
  --style content/global/style.yaml \
  --output slide-deck/presentation.pptx \
  --allow-scripts

When validation fails, the build raises ContentExtraError with a message identifying the violation and file path.

Script Reference

All operations are available through the PowerShell orchestrator (Invoke-PptxPipeline.ps1) or directly via the Python scripts. The PowerShell script manages the Python virtual environment and dependency installation automatically via uv sync.

Build a Slide Deck

./scripts/Invoke-PptxPipeline.ps1 -Action Build `
  -ContentDir content/ `
  -StylePath content/global/style.yaml `
  -OutputPath slide-deck/presentation.pptx

python scripts/build_deck.py \
  --content-dir content/ \
  --style content/global/style.yaml \
  --output slide-deck/presentation.pptx

Reads all content/slide-*/content.yaml files in numeric order and generates the complete deck. Executes content-extra.py files when present.

Build from a Template

> [!WARNING] > --template creates a NEW presentation inheriting only slide masters, layouts, and theme from the template. All existing slides are discarded. Use --source for partial rebuilds.

./scripts/Invoke-PptxPipeline.ps1 -Action Build `
  -ContentDir content/ `
  -StylePath content/global/style.yaml `
  -OutputPath slide-deck/presentation.pptx `
  -TemplatePath corporate-template.pptx

python scripts/build_deck.py \
  --content-dir content/ \
  --style content/global/style.yaml \
  --output slide-deck/presentation.pptx \
  --template corporate-template.pptx

Loads slide masters and layouts from the template PPTX. Layout names in each slide's content.yaml resolve against the template's layouts, with optional name mapping via the layouts section in style.yaml. Populate themed layout placeholders using the placeholders section in content YAML.

Update Specific Slides

> [!IMPORTANT] > Use --source (not --template) for partial rebuilds. Combining --template and --source is not supported.

./scripts/Invoke-PptxPipeline.ps1 -Action Build `
  -ContentDir content/ `
  -StylePath content/global/style.yaml `
  -OutputPath slide-deck/presentation.pptx `
  -SourcePath slide-deck/presentation.pptx `
  -Slides "3,7,15"

python scripts/build_deck.py \
  --content-dir content/ \
  --style content/global/style.yaml \
  --source slide-deck/presentation.pptx \
  --output slide-deck/presentation.pptx \
  --slides 3,7,15

Opens the existing deck, clears shapes on the specified slides, rebuilds them in-place from their content.yaml, and saves. All other slides remain untouched. After building, verify the output slide count matches the original deck.

Extract Content from Existing PPTX

./scripts/Invoke-PptxPipeline.ps1 -Action Extract `
  -InputPath existing-deck.pptx `
  -OutputDir content/

python scripts/extract_content.py \
  --input existing-deck.pptx \
  --output-dir content/

Extracts text, shapes, images, and styling from an existing PPTX into the content/ folder structure. Creates content.yaml files for each slide and populates the global/style.yaml from detected patterns.

Extract Specific Slides

./scripts/Invoke-PptxPipeline.ps1 -Action Extract `
  -InputPath existing-deck.pptx `
  -OutputDir content/ `
  -Slides "3,7,15"

python scripts/extract_content.py \
  --input existing-deck.pptx \
  --output-dir content/ \
  --slides 3,7,15

Extracts only the specified slides (plus the global style). Useful for targeted updates on large decks.

Extraction Limitations

• Picture shapes that reference external (linked) images instead of embedded blobs are recorded with path: LINKED_IMAGE_NOT_EMBEDDED. The script does not crash but the image must be re-embedded manually.
• When text elements inherit font, size, or color from the slide master or layout, the extraction records no inline styling. Content YAML for these elements needs explicit font properties added before rebuild.
• The detect_global_style() function uses frequency analysis across all slides. For decks with mixed styling, review and adjust style.yaml values manually after extraction.

Validate a Deck

./scripts/Invoke-PptxPipeline.ps1 -Action Validate `
  -InputPath slide-deck/presentation.pptx `
  -ContentDir content/

The Validate action runs a two- or three-step pipeline:

1. Export — Clears stale slide images from the output directory, then renders slides to JPG images via LibreOffice (PPTX → PDF → JPG). When -Slides is used, output images are named to match original slide numbers (e.g., slide-023.jpg for slide 23), not sequential PDF page numbers.
2. PPTX validation — Checks PPTX-only properties (validate_deck.py) for speaker notes and slide count.
3. Vision validation (optional) — Sends slide images to a vision-capable model via the Copilot SDK (validate_slides.py) for visual quality checks. Runs when -ValidationPrompt or -ValidationPromptFile is provided.

For validation criteria (element positioning, visual quality, color contrast, content completeness), see pptx.instructions.md Validation Criteria.

Built-in System Message

The validate_slides.py script includes a built-in system message that focuses on issue detection only (not full slide description). It checks overlapping elements, text overflow/cutoff, decorative line mismatch after title wraps, citation/footer collisions, tight spacing, uneven gaps, insufficient edge margins, alignment inconsistencies, low contrast, narrow text boxes, and leftover placeholders. For dense slides, near-edge placement or tight boundaries are acceptable when readability is not materially affected. The -ValidationPrompt parameter provides supplementary user-level context and does not need to repeat these checks.

Validate with Vision Checks

./scripts/Invoke-PptxPipeline.ps1 -Action Validate `
  -InputPath slide-deck/presentation.pptx `
  -ContentDir content/ `
  -ValidationPrompt "Validate visual quality. Focus on recently modified slides for content accuracy." `
  -ValidationModel claude-haiku-4.5

Vision validation results are written to validation-results.json in the image output directory, containing raw model responses per slide with quality findings. Per-slide response text is also written to slide-NNN-validation.txt files next to each slide image.

Validate Specific Slides

./scripts/Invoke-PptxPipeline.ps1 -Action Validate `
  -InputPath slide-deck/presentation.pptx `
  -ContentDir content/ `
  -Slides "3,7,15"

Validates only the specified slides. When content directories cover fewer slides than the PPTX, the slide count check reports an informational note rather than an error.

validate_slides.py CLI Reference

| Flag | Required | Default | Description | |-------------------|-------------------------------------|--------------------|-----------------------------------------------| | --image-dir | Yes | — | Directory containing slide-NNN.jpg images | | --prompt | One of --prompt / --prompt-file | — | Validation prompt text | | --prompt-file | One of --prompt / --prompt-file | — | Path to file containing the validation prompt | | --model | No | claude-haiku-4.5 | Vision model ID | | --output | No | stdout | JSON results file path | | --slides | No | all | Comma-separated slide numbers to validate | | -v, --verbose | No | — | Enable debug-level logging |

validate_deck.py CLI Reference

| Flag | Required | Default | Description | |-------------------|----------|---------|-----------------------------------------------------------------------| | --input | Yes | — | Input PPTX file path | | --content-dir | No | — | Content directory for slide count comparison | | --slides | No | all | Comma-separated slide numbers to validate | | --output | No | stdout | JSON results file path | | --report | No | — | Markdown report file path | | --per-slide-dir | No | — | Directory for per-slide JSON files (slide-NNN-deck-validation.json) |

Validation Outputs

When run through the pipeline, validation produces these files in the image output directory:

| File | Format | Content | |-

…

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: microsoft
• Source: microsoft/hve-core
• License: MIT

Install and usage instructions live in the source repository linked above.