AgentiaPT/vela-slides / skills/vela-slides
vela-slides
UnverifiedCreate presentation decks using the Vela engine. Compact DSL format — never verbose JSON. Also loads, extracts, and edits existing decks.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 6.5 KB |
| app/vela.jsx | other | 2.3 MB |
| examples/vela-demo.json | config | 87.0 KB |
| references/block-schema.md | markdown | 25.6 KB |
| references/design-patterns.md | markdown | 20.5 KB |
| references/formats.md | markdown | 4.6 KB |
| references/themes.md | markdown | 4.5 KB |
| scripts/_safe_term.py | script | 10.2 KB |
| scripts/assemble.py | script | 9.3 KB |
| scripts/validate.py | script | 16.7 KB |
| scripts/vela.py | script | 115.0 KB |
references/design-patterns.md
View on GitHub# Vela Design Patterns & Slide Archetypes
> **v12.5** · 2026-03-23
## Slide Archetypes
Every slide should follow one of these proven patterns. Mix them throughout a deck for visual variety.
### 1. Title Slide
Opening slide — centered, gradient bg, badge + 4xl heading + subtitle + presenter line. `verticalAlign: "center"`, `align: "center"`. Duration: 20s.
### 2. Section Break
Divides major sections — centered xl icon in circle + 3xl heading + subtitle. High-contrast gradient. Duration: 15s.
### 3. Key Point + Supporting Detail
Workhorse content slide — badge label → spacer → 2xl assertion heading with icon → body text → callout. Duration: 60-90s.
### 4. Feature Grid
2-3 col grid with icon + heading + text per cell. Each cell gets subtle colored background (`rgba(...,0.08)`) + border. Duration: 60-90s.
### 5. Metrics / Stats
3-col grid of metric blocks with large values (`3xl`), icons, and ALL-CAPS labels. Gradient bg. Duration: 20-40s.
### 6. Process / Flow
Horizontal flow block with 3-5 icon steps + sublabels connected by arrows. Add `loop: true` for iterative processes (agent loops, ReAct, OODA, feedback cycles). Add `gate: true` on items for human-review checkpoints. Add italic footnote below. Duration: 90-120s.
### 7. Comparison Table
Table block with striped rows, colored headers. Heading with icon above. Duration: 60-120s.
### 8. Timeline / Roadmap
Horizontal timeline block with 3-5 dated milestones. Duration: 60s.
### 9. Closing / CTA
Centered layout — xl icon + 3xl heading + body text + tag-group for contact/links. Gradient bg. Duration: 30s.
### 10. Layer Diagram *(v13)*
Stacked architecture / composition visual using 1-col grid with `direction: "row"` cells.
**When to use:** Architecture stacks, OSI model, testing pyramid, context window composition, agent anatomy, MCP layers, any layered system where order implies hierarchy.
**Design rules:**
- Use `cols: 1` grid — each cell = one horizontal layer bar
- Each cell: `direction: "row"` with `icon` + `heading` + `text` flowing inline
- Left border color = layer's semantic color
- Background opacity: `0.06–0.10` (subtle, not loud)
- Tight padding: `10px 16px` to `14px 20px`
- Keep sublabel text as plain `text` block, not badges (cleaner)
```json
{
"bg": "#0f172a",
"color": "#e2e8f0",
"accent": "#3b82f6",
"padding": "36px 48px",
"duration": 90,
"blocks": [
{ "type": "heading", "text": "Composable Layers", "size": "2xl" },
{ "type": "text", "text": "Spec tells the loop WHAT. Execution does IT. CI validates EACH commit.", "size": "md", "color": "#94a3b8" },
{ "type": "spacer", "h": 12 },
{
"type": "grid", "cols": 1, "gap": 8,
"items": [
{
"direction": "row",
"blocks": [
{ "type": "icon", "name": "FileText", "size": "md", "color": "#3b82f6", "circle": true, "bg": "#3b82f620" },
{ "type": "heading", "text": "Specification", "size": "lg" },
{ "type": "text", "text": "Intent & constraints — the WHAT and WHY", "size": "sm", "color": "#94a3b8" }
],
"style": { "background": "rgba(59,130,246,0.08)", "borderLeft": "4px solid #3b82f6", "borderRadius": "8px", "padding": "12px 18px" }
},
{
"direction": "row",
"blocks": [
{ "type": "icon", "name": "Play", "size": "md", "color": "#8b5cf6", "circle": true, "bg": "#8b5cf620" },
{ "type": "heading", "text": "Execution", "size": "lg" },
{ "type": "text", "text": "Ralph loop / HITL / Parallel sub-agents", "size": "sm", "color": "#94a3b8" }
],
"style": { "background": "rgba(139,92,246,0.08)", "borderLeft": "4px solid #8b5cf6", "borderRadius": "8px", "padding": "12px 18px" }
},
{
"direction": "row",
"blocks": [
{ "type": "icon", "name": "CheckCircle", "size": "md", "color": "#34d399", "circle": true, "bg": "#34d39920" },
{ "type": "heading", "text": "Verification", "size": "lg" },
{ "type": "text", "text": "CI / Tests / Evals — per commit", "size": "sm", "color": "#94a3b8" }
],
"style": { "background": "rgba(52,211,153,0.08)", "borderLeft": "4px solid #34d399", "borderRadius": "8px", "padding": "12px 18px" }
}
]
}
]
}
```
### 11. Gated Pipeline *(v13)*
Workflow with human-review checkpoints between stages.
**When to use:** CI/CD approval stages, spec-driven development phases, deployment pipelines, compliance workflows, code review gates.
**Design rules:**
- Use `flow` with `gate: true` on items that need a checkpoint AFTER them
- Set `gateIcon`, `gateLabel`, `gateColor` at block level (shared across all gates)
- Gates render as dashed-circle checkpoints aligned to icon centerline
- Amber/gold (`#f59e0b`) is the natural gate color — implies "pause and review"
- Keep flow to 3-5 items max (gates add visual width)
```json
{
"bg": "#0f172a",
"color": "#e2e8f0",
"accent": "#3b82f6",
"padding": "36px 48px",
"duration": 120,
"blocks": [
{ "type": "heading", "text": "CI/CD with Approval Gates", "size": "2xl", "icon": "GitBranch" },
{ "type": "spacer", "h": 20 },
{
"type": "flow",
"items": [
{ "icon": "Code", "label": "Build", "sublabel": "Compile + lint" },
{ "icon": "FlaskConical", "label": "Test", "sublabel": "Unit + E2E", "gate": true },
{ "icon": "Eye", "label": "Staging", "sublabel": "Preview deploy", "gate": true },
{ "icon": "Rocket", "label": "Production", "sublabel": "Live" }
],
"gateIcon": "UserCheck",
"gateLabel": "Approve",
"gateColor": "#f59e0b",
"arrowColor": "#3b82f6",
"direction": "horizontal"
}
]
}
```
### 12. Spectrum / Continuum *(v13)*
Position something between two extremes using a progress bar with endpoint labels.
**When to use:** Methodology positioning, risk scales, maturity models, capability continuum, team skill assessment, cost-quality tradeoff.
**Design rules:**
- Single `progress` item at a specific `value` (0-100) position
- `leftLabel` / `rightLabel` define the extremes
- Optional `leftIcon` / `rightIcon` reinforce the endpoints
- `showValue: false` — the position tells the story, not the number
- `annotation` below explains positioning in context
- Use `height: 10-12` for visual weight
- Can stack multiple progress blocks for multi-dimension spectrums
```json
{
"bg": "#0f172a",
"color": "#e2e8f0",
"accent": "#3b82f6",
"padding": "36px 48px",
"duration": 60,
"blocks": [
{ "type": "heading", "text": "Not Waterfall, Not Vibe Coding", "size": "2xl" },
{ "type": "text", "text": "SDD sits in the disciplined middle.", "size": "md", "color": "#94a3b8" },
{ "type": "spacer", "h": 16 },
{
"type": "progress",
"items": [{ "label": "SDD", "value": 35, "color": "#3b82f6" }],
"leftLabel": "Waterfall", "rightLabel": "Vibe Coding",
"leftIcon": "FileText", "rightIcon": "Zap",
"showValue": false, "trackColor": "#1e293b", "height": 10,
"annotation": "Shorter feedback loops than waterfall, better output than vibe coding",
"annotationColor": "#94a3b8"
}
]
}
```
### 13. Diagram Slide *(v14)*
Custom SVG diagram for non-linear visuals that structured blocks can't express.
**When to use:** Architecture diagrams with feedback loops, fan-out patterns (1→N), mesh connectors (M×N), probability distributions, custom layouts with arrows between arbitrary elements.
**Design rules:**
- Badge label → spacer → `svg` block → optional callout or text below
- Always use theme tokens (`{{accent}}`, `{{color}}`, `{{muted}}`, `{{bg}}`) in SVG markup
- Keep `viewBox` height ≤ 300px
- Use `monospace` font-family for text in SVGs
- Prefer stroke-based outlines over filled shapes
- Set `maxWidth: "90%"` and `align: "center"` for typical diagrams
- Use `caption` for diagram labels instead of separate text blocks
```json
{
"bg": "#0f172a",
"color": "#e2e8f0",
"accent": "#3b82f6",
"padding": "36px 48px",
"duration": 90,
"blocks": [
{ "type": "badge", "text": "ARCHITECTURE", "icon": "Layers", "bg": "#3b82f620", "color": "#60a5fa" },
{ "type": "spacer", "h": 8 },
{ "type": "heading", "text": "Agent Core with Context Fan-Out", "size": "2xl", "icon": "Brain" },
{ "type": "spacer", "h": 12 },
{
"type": "svg",
"maxWidth": "90%",
"align": "center",
"markup": "<svg viewBox='0 0 600 200' xmlns='http://www.w3.org/2000/svg'><rect x='220' y='60' width='160' height='80' rx='12' fill='none' stroke='{{accent}}' stroke-width='2'/><text x='300' y='105' text-anchor='middle' fill='{{color}}' font-size='16' font-weight='bold' font-family='monospace'>Agent Core</text><rect x='10' y='20' width='120' height='40' rx='6' fill='none' stroke='{{accent}}' stroke-width='1.5'/><text x='70' y='45' text-anchor='middle' fill='{{muted}}' font-size='11' font-family='monospace'>RAG Store</text><rect x='10' y='80' width='120' height='40' rx='6' fill='none' stroke='{{accent}}' stroke-width='1.5'/><text x='70' y='105' text-anchor='middle' fill='{{muted}}' font-size='11' font-family='monospace'>Memory</text><rect x='10' y='140' width='120' height='40' rx='6' fill='none' stroke='{{accent}}' stroke-width='1.5'/><text x='70' y='165' text-anchor='middle' fill='{{muted}}' font-size='11' font-family='monospace'>Profile</text><line x1='130' y1='40' x2='220' y2='80' stroke='{{muted}}' stroke-width='1'/><line x1='130' y1='100' x2='220' y2='100' stroke='{{muted}}' stroke-width='1'/><line x1='130' y1='160' x2='220' y2='120' stroke='{{muted}}' stroke-width='1'/><rect x='470' y='20' width='120' height='40' rx='6' fill='none' stroke='#f59e0b' stroke-width='1.5'/><text x='530' y='45' text-anchor='middle' fill='#fcd34d' font-size='11' font-family='monospace'>Tool A</text><rect x='470' y='80' width='120' height='40' rx='6' fill='none' stroke='#f59e0b' stroke-width='1.5'/><text x='530' y='105' text-anchor='middle' fill='#fcd34d' font-size='11' font-family='monospace'>Tool B</text><rect x='470' y='140' width='120' height='40' rx='6' fill='none' stroke='#ef4444' stroke-width='1.5'/><text x='530' y='165' text-anchor='middle' fill='#fca5a5' font-size='11' font-family='monospace'>Output</text><line x1='380' y1='80' x2='470' y2='40' stroke='#f59e0b' stroke-width='1'/><line x1='380' y1='100' x2='470' y2='100' stroke='#f59e0b' stroke-width='1'/><line x1='380' y1='120' x2='470' y2='160' stroke='#ef4444' stroke-width='1'/></svg>",
"caption": "Context fan-in → Agent reasoning → Tool fan-out"
}
]
}
```
### 14. Loop Flow *(v14)*
Iterative process diagram — flow block with return arrow.
**When to use:** ReAct agent loops, OODA loops, Reflexion patterns, feedback cycles, retry mechanisms, any iterative process where the last step feeds back to the first.
**Design rules:**
- Use `flow` with `loop: true` (NOT `svg` — flow handles this natively now)
- Set `loopLabel` to describe the iteration condition
- `loopStyle: "dashed"` (default) is best for most cases
- Keep to 3-4 items — loops already add visual complexity
- Combine with `gate: true` for human-in-the-loop review points
```json
{
"bg": "#0f172a",
"color": "#e2e8f0",
"accent": "#10b981",
"padding": "36px 48px",
"duration": 90,
"blocks": [
{ "type": "badge", "text": "AGENT PATTERN", "icon": "Bot", "bg": "#10b98120", "color": "#34d399" },
{ "type": "spacer", "h": 8 },
{ "type": "heading", "text": "ReAct Loop", "size": "2xl", "icon": "RefreshCw" },
{ "type": "spacer", "h": 16 },
{
"type": "flow",
"items": [
{ "icon": "Brain", "label": "Thought", "sublabel": "Reason about task" },
{ "icon": "Zap", "label": "Action", "sublabel": "Call tool / API" },
{ "icon": "Eye", "label": "Observe", "sublabel": "Read result" }
],
"loop": true,
"loopLabel": "repeat until task complete",
"loopColor": "#64748b",
"loopStyle": "dashed",
"arrowColor": "#10b981"
},
{ "type": "spacer", "h": 12 },
{ "type": "callout", "text": "The loop makes reasoning visible — each iteration is an auditable step.", "icon": "Shield", "bg": "rgba(16,185,129,0.1)", "border": "#10b981", "color": "#e2e8f0" }
]
}
```
---
### 15. Annotated Study Slide *(v12.32)*
Any archetype above **plus** an offline `studyNotes` block that renders in the 🎓 student panel with zero API calls.
**When to use:** course material, self-study decks, accessible presentations, and any environment where a live AI tutor can't be guaranteed (shared files, non-Claude hosts, offline viewing).
**Design rules:**
- Keep `studyNotes.text` tight — 150–400 words (3–6 paragraphs). The panel is a companion, not a handout.
- Include a `diagram` only when the visual genuinely adds information. It's not decoration.
- Write `questions` that probe *why* and *how*, not *what* (the slide already answered *what*).
- Populate `glossary` with the 3–8 key jargon words that actually appear on the slide. Don't catalogue every noun.
- Inline X-Ray links: `[term](#key)` where keys are lowercase. Unknown keys render as plain label text — safe fallback.
- For a definition + "learn more" combo, prefer `glossary[term] = { definition, url }` over raw inline `[label](https://…)` — it keeps the prose clean and surfaces the definition as a popover.
- Questions become clickable Vera prompts only when the live API is reachable. Design them so they still make sense as static "questions to ponder" when read offline.
**Minimal example (compact format, inside a slide):**
```json
"sN": {
"text": "Agents close the loop: **plan → act → observe → revise**. Each iteration is an auditable step, unlike a single-shot LLM call. See the [ReAct paper](https://arxiv.org/abs/2210.03629) for the original formulation of this pattern, or explore [what an agent really is](#agent).",
"questions": [
"Why does exposing the loop matter for trust and debugging?",
"When would you stop iterating — cost, confidence, or task completion?"
],
"glossary": {
"agent": { "definition": "A goal-driven loop that plans, acts, and observes — not just a single prompt.", "url": "https://example.com/agents" }
}
}
```
The 🎓 marker appears in the TOC, gallery thumbnails, and the slide viewer automatically. No extra wiring needed.
---
### Spacing & Rhythm
- Start content slides with a `badge` (section label) → `spacer h:8` → `heading`
- Use `spacer h:12-16` between heading and content
- Use `spacer h:16-24` between content sections
- Never stack two headings without a spacer
### Visual Variety Checklist
For a 10-slide deck, aim for at least:
- 1 title slide (archetype 1)
- 1-2 section breaks (archetype 2)
- 2-3 content slides with different block types
- 1 metrics slide (archetype 5)
- 1 flow, loop flow, gated pipeline, or timeline (archetype 6, 14, 11, or 8)
- 1 grid, layer diagram, or comparison (archetype 4, 10, or 7)
- 1 spectrum if positioning is relevant (archetype 12)
- 1 diagram if non-linear visuals needed (archetype 13)
- 1 closing slide (archetype 9)
### Anti-Patterns (Avoid These)
- ❌ Heading + bullets on every slide (monotonous)
- ❌ More than 7 blocks per slide (overflow)
- ❌ No icons anywhere (bland)
- ❌ Identical bg on all slides (flat)
- ❌ Missing duration (breaks timing)
- ❌ Body text at `sm` or below (use `md` minimum for readable content, `lg` preferred)
- ❌ Using `xs` for anything except monospace labels or timestamps
- ❌ Grid with more than 3 columns (too cramped at 960px)
- ❌ Bullets with 8+ items (split into two slides)
- ❌ Label headlines ("Results") instead of assertion headlines ("Revenue Up 30%")
- ❌ Layer diagram cells with badges inside rows (use plain text sublabels)
- ❌ Layer diagram backgrounds above 0.12 opacity (too loud)
- ❌ Spectrum progress bars with `showValue: true` (position tells the story)
- ❌ Using `svg` block when `flow` with `loop: true` would work (use structured blocks first)
- ❌ SVG markup without `viewBox` attribute (breaks responsiveness)
- ❌ SVG with hardcoded colors instead of theme tokens (breaks light/dark switching)
- ❌ SVG `viewBox` height > 300px (dominates the slide)
- ❌ Slide-level `align: "center"` with `svg`, `divider`, or `progress` blocks (they collapse to zero width — use `align: "left"` + per-block centering instead)
### Content Density Guidelines
| Slide Type | Max Blocks | Max Text Lines | Target Duration |
|------------|-----------|----------------|-----------------|
| Title | 4-5 | 3 | 15-30s |
| Section Break | 3-4 | 2 | 10-15s |
| Key Point | 5-6 | 6-8 | 60-90s |
| Feature Grid | 4-5 | 9-12 (across cells) | 60-90s |
| Metrics | 4-5 | 3-4 | 20-40s |
| Process/Flow | 4-5 | 5-7 | 90-120s |
| Loop Flow | 4-6 | 5-7 | 90-120s |
| Layer Diagram | 4-6 | 6-9 (across layers) | 60-90s |
| Gated Pipeline | 4-5 | 5-7 | 90-120s |
| Spectrum | 4-6 | 4-6 | 45-90s |
| Diagram (svg) | 3-5 | 3-4 + SVG labels | 60-90s |
| Table | 3-4 | varies | 60-120s |
| Code | 3-4 | 15 lines max | 90-180s |
| Closing | 4-5 | 3 | 20-30s |
---
### Known Issues & Gotchas
#### ⚠️ `align: "center"` slides shrink-wrap blocks — SVGs and dividers collapse
**Problem:** When a slide has `align: "center"`, Vela shrink-wraps all blocks to their intrinsic content width. Blocks with no intrinsic width — `divider` (a border line), `svg` (responsive viewBox), and `progress` — collapse to zero or near-zero width and become invisible.
**Symptoms:**
- `divider` blocks disappear entirely
- `svg` blocks don't render (no width to fill)
- Layout looks like those blocks were never added
**Fix:** Use `align: "left"` at slide level and center content via individual block properties:
- Text/heading blocks: add `"align": "center"` per block
- SVG blocks: add `"align": "center"` + `"maxWidth": "60%"` (SVG renderer uses `margin: 0 auto`)
- For visual separators on centered slides: use an `svg` block with an explicit rect instead of `divider`
**Example — centered closing slide that works:**
```json
{
"align": "left",
"verticalAlign": "center",
"blocks": [
{ "type": "svg", "markup": "<svg viewBox='0 0 420 70' .../>", "maxWidth": "60%", "align": "center" },
{ "type": "text", "text": "Subtitle", "align": "center" },
{ "type": "svg", "markup": "<svg viewBox='0 0 200 4'><rect ... fill='#34D399'/></svg>", "maxWidth": "25%", "align": "center" },
{ "type": "heading", "text": "Contact Name", "align": "center" }
]
}
```
**Rule of thumb:** Never use slide-level `align: "center"` if the slide contains `svg`, `divider`, or `progress` blocks. Use `align: "left"` + per-block centering instead.
---
## Cols Layout Patterns
The `layout: "cols"` arranges blocks into two columns using `L` (left) and `R` (right) arrays. An optional `blocks` (compact: `B`) array renders full-width above the columns as a header area. `contentFlex` / `imageFlex` control column widths (default 1:1). `splitGap` controls the gap between columns (default 32px).
### Header + Balanced Cols (most common)
B: badge + heading + context callout
L: primary content (steps, bullets, table)
R: supporting visual (SVG, metrics, flow diagram)
```json
{
"layout": "cols",
"contentFlex": 3,
"imageFlex": 2,
"blocks": [
{ "type": "badge", "text": "ARCHITECTURE" },
{ "type": "heading", "text": "System Overview", "size": "2xl" }
],
"L": [
{ "type": "bullets", "items": ["Microservices", "Event-driven", "Auto-scaling"] }
],
"R": [
{ "type": "flow", "items": [{"icon": "Upload", "label": "In"}, {"icon": "Cpu", "label": "Process"}, {"icon": "Download", "label": "Out"}] }
]
}
```
### Pure Side-by-Side (no header)
B: (empty / omitted)
L: "Before" content with badge
R: "After" content with badge
```json
{
"layout": "cols",
"L": [
{ "type": "badge", "text": "BEFORE", "bg": "#dc2626" },
{ "type": "bullets", "items": ["Manual deploys", "3-hour rollbacks"] }
],
"R": [
{ "type": "badge", "text": "AFTER", "bg": "#16a34a" },
{ "type": "bullets", "items": ["CI/CD pipeline", "5-min rollbacks"] }
]
}
```
### Content + Sidebar
contentFlex: 4, imageFlex: 1
L: main content
R: 1-2 metrics or single visual
```json
{
"layout": "cols",
"contentFlex": 4,
"imageFlex": 1,
"blocks": [
{ "type": "heading", "text": "Q4 Results", "size": "2xl" }
],
"L": [
{ "type": "steps", "items": [{"title": "Step 1", "text": "Launched beta"}, {"title": "Step 2", "text": "Scaled to 10K users"}] }
],
"R": [
{ "type": "metric", "value": "98%", "label": "Uptime", "size": "2xl" }
]
}
```
### Common Ratio Patterns
| Ratio | contentFlex:imageFlex | Use case |
|-------|----------------------|----------|
| Equal | 1:1 | Before/after, comparison |
| Wide left | 3:2 | Text + supporting visual |
| Wide right | 2:3 | Summary + detailed steps |
| Sidebar | 4:1 or 1:4 | Main content + narrow metric strip |
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| canvas-designanthropics/skillsCreate beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| docxanthropics/skillsUse this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files) or Word templates (.dotx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', '.dotx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx or .dotx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| pdfanthropics/skillsUse this skill whenever the user wants to do anything with PDF files. This includes reading or extracting text/tables from PDFs, combining or merging multiple PDFs into one, splitting PDFs apart, rotating pages, adding watermarks, creating new PDFs, filling PDF forms, encrypting/decrypting PDFs, extracting images, and OCR on scanned PDFs to make them searchable. If the user mentions a .pdf file or asks to produce one, use this skill. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| pptxanthropics/skillsUse this skill any time a .pptx or .potx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx or .potx file (even if the extracted content will be used elsewhere, like in an email or summary); editing, modifying, or updating existing presentations; combining or splitting slide files; working with templates (.potx), layouts, speaker notes, or comments. Trigger whenever the user mentions "deck," "slides," "presentation," or references a .pptx or .potx filename, regardless of what they plan to do with the content afterward. If a .pptx or .potx file needs to be opened, created, or touched, use this skill. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |