utility-slideshow-creator
About
This skill generates professional PowerPoint presentations (.pptx/.pdf) from JSON specifications, automatically selecting from 18 slide types with dark/light variants and handling content-to-layout logic. It includes a default theme and supports custom theming, making it ideal for creating stakeholder updates or product reviews programmatically. Developers can use it to produce presentation decks without manual design decisions at generation time.
Quick Install
Claude Code
Recommendednpx skills add product-on-purpose/pm-skills -a claude-code/plugin add https://github.com/product-on-purpose/pm-skillsgit clone https://github.com/product-on-purpose/pm-skills.git ~/.claude/skills/utility-slideshow-creatorCopy and paste this command in Claude Code to install this skill
Documentation
Slideshow Creator
Generate professional presentations (.pptx and .pdf) from a JSON deck specification. Zero design decisions at generation time . Claude selects slide types and fills content slots; all visual properties are pre-decided by the theme.
When to Use
- Creating slide decks for stakeholder updates, product reviews, team presentations
- Generating professional .pptx files from content briefs
- Producing consistent presentations without manual formatting
- Exporting presentations to PDF for sharing or archiving
When NOT to Use
- Creating complex data visualizations (use dedicated charting tools)
- Building interactive web presentations (this produces .pptx, not HTML)
- Editing existing PowerPoint files (this creates new decks from scratch)
How It Works
Two-phase architecture:
- Spec phase (Claude) . Read the content brief, select slide types using decision logic, write a JSON deck specification. Content slots have character limits calibrated to prevent overflow. This is the only phase that costs tokens.
- Generation phase (Local script) . Run
node scripts/generate-deck.js deck-spec.jsonto produce .pptx. Optionally runnode scripts/export-pdf.mjs deck-spec.jsonfor PDF. Deterministic rendering, zero token cost.
Both outputs come from the same JSON spec, so .pptx and .pdf always match.
18 Slide Types
| # | Type Key | Purpose | Default Variant |
|---|---|---|---|
| 1 | title_dark | Opening slide (bold) | dark only |
| 2 | title_light | Opening slide (internal/lighter) | light only |
| 3 | section | Section divider | dark |
| 4 | content | Paragraph explanation | light |
| 5 | bullets | 3-6 key points | light |
| 6 | two_col | Side-by-side comparison | light |
| 7 | stat | Single key metric | light |
| 8 | dual_stat | Two metrics compared | dark |
| 9 | quote | Testimonial or pull quote | dark |
| 10 | three_card | Three parallel concepts | dark |
| 11 | four_grid | Four concepts in 2x2 grid | dark |
| 12 | timeline | Dates or milestones (max 6) | light |
| 13 | process_flow | Sequential workflow (max 5) | light |
| 14 | agenda | Meeting agenda (max 7) | light |
| 15 | highlight | Key finding or executive summary | light |
| 16 | table | Tabular data | light |
| 17 | icon_rows | Feature list with markers (max 4) | light |
| 18 | closing | End slide | dark only |
Full slot definitions and character limits: references/slide-types.md
Decision Logic (Quick Reference)
| Content Pattern | Use |
|---|---|
| Opening the deck (bold) | title_dark |
| Opening the deck (internal) | title_light |
| Transitioning between topics | section |
| Paragraph explanation | content |
| List of 3-6 points | bullets |
| Side-by-side comparison | two_col |
| Single key metric | stat |
| Two metrics compared | dual_stat |
| Testimonial or pull quote | quote |
| Three parallel concepts | three_card |
| Four concepts in a grid | four_grid |
| Dates or milestones | timeline |
| Sequential workflow | process_flow |
| Meeting agenda | agenda |
| Key finding or summary | highlight |
| Tabular data | table |
| Feature list with markers | icon_rows |
| Ending the deck | closing |
Full decision logic with variant strategy: references/decision-logic.md
JSON Deck Spec Format
{
"title": "Q3 Product Update",
"author": "Product Team",
"footerText": "Internal . Q3 Review",
"slides": [
{ "type": "title_dark", "title": "Q3 Product Update", "subtitle": "October 2026" },
{ "type": "stat", "stat": "94%", "label": "Customer satisfaction score", "accentColor": "secondary" },
{ "type": "bullets", "title": "What Shipped", "bullets": ["Feature A", "Feature B", "Feature C"] },
{ "type": "closing" }
]
}
Colors accept theme token names ("accent", "secondary", "tertiary", "warm") or 6-character hex strings ("2563EB").
Full schema and workflow: references/TEMPLATE.md and references/platform-rules.md
Instructions
- Read the content brief . Understand topic, audience, length, specific requirements
- Plan the deck . Select slide types using the decision logic table. Assign dark/light variants for visual rhythm (alternate to avoid monotony).
- Write the JSON deck specification . Fill content slots, respecting character limits from
references/slide-types.md - Run the generation script .
node scripts/generate-deck.js deck-spec.json - Optionally export PDF .
node scripts/export-pdf.mjs deck-spec.json output.pdf - Report the output . Tell the user where the file(s) are
Output Contract
- Planning artifact: JSON deck specification (the file Claude writes)
- Final output: .pptx file (+ optional .pdf), generated locally from the spec
- Quality gate: All items in the quality checklist pass
Quality Checklist
- Every slide has a valid type key (one of the 18 defined types)
- All content slots respect character limits from
references/slide-types.md - Dark/light variants alternate for visual rhythm (no 3+ consecutive same-variant slides)
- Deck has a title slide (first) and a closing slide (last)
- Section dividers separate distinct topics in decks > 6 slides
- Speaker notes included for key slides (stats, highlights, transitions)
- JSON is valid (no trailing commas, proper quoting, correct nesting)
- Color values are valid theme tokens or 6-character hex strings (no
#prefix)
References
| File | Purpose |
|---|---|
references/TEMPLATE.md | JSON deck specification template with field documentation |
references/EXAMPLE.md | Worked example: Q3 Product Update deck (9 slides) |
references/slide-types.md | All 18 slide types: content slots, character limits, variants |
references/decision-logic.md | Content pattern → slide type mapping, variant strategy, deck sizing |
references/platform-rules.md | pptxgenjs requirements, Google Slides compatibility, output format |
GitHub Repository
Related Skills
content-collections
MetaThis skill provides a production-tested setup for Content Collections, a TypeScript-first tool that transforms Markdown/MDX files into type-safe data collections with Zod validation. Use it when building blogs, documentation sites, or content-heavy Vite + React applications to ensure type safety and automatic content validation. It covers everything from Vite plugin configuration and MDX compilation to deployment optimization and schema validation.
polymarket
MetaThis skill enables developers to build applications with the Polymarket prediction markets platform, including API integration for trading and market data. It also provides real-time data streaming via WebSocket to monitor live trades and market activity. Use it for implementing trading strategies or creating tools that process live market updates.
creating-opencode-plugins
MetaThis skill helps developers create OpenCode plugins that hook into 25+ event types like commands, files, and LSP operations. It provides the plugin structure, event API specifications, and implementation patterns for JavaScript/TypeScript modules. Use it when you need to intercept, monitor, or extend the OpenCode AI assistant's lifecycle with custom event-driven logic.
sglang
MetaSGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.
