---
name: actcenter_cowork-to-design-export
description: Export any material produced inside Claude/Cowork into a design-ready format — PNG, SVG, PDF, PPTX, DOCX, Figma-ready frames, Canva-ready assets, social-media-ready crops, transparent backgrounds, brand-aware versions. Use when the user says "export this to design", "send this to Figma/Canva", "convert this report to a deck", "make this LinkedIn-ready", "preciso disso em PNG transparente", "vira slide", "exporta pro Canva", "social post a partir desse texto", or when a deliverable needs to leave the Claude session and enter a designer's workflow. Bridges Claude outputs and design tools.
metadata:
  type: skill
  version: 1.0
  owner: actcenter
---

# actcenter_cowork-to-design-export

Master skill for the "now I need this in design" handoff. Takes any artifact produced in a Claude session — a markdown report, a table, a chart, a flow, a copy block, a SVG diagram — and packages it for the next tool: Figma, Canva, Adobe, PowerPoint, social media, blog publishing.

## When to use

Trigger on any of:

- "Export this to design"
- "I need this in Figma / Canva"
- "Make this a slide / deck / one-pager"
- "Turn this into a LinkedIn post / Instagram graphic / Twitter image"
- "Give me this as transparent PNG"
- "Vetorize esse gráfico", "exporta em SVG", "PNG 2x"
- "I want to drop this into my brand template"
- A markdown table or text block that the user clearly wants visual

## What this skill does

Three steps, every time:

### Step 1 — Identify the source

Read what's in the conversation, the artifact, or the file the user pointed at. Categorize it:

- **Prose / report** → DOCX or PDF
- **Data table** → XLSX, CSV, or styled PNG/SVG
- **Diagram / flow** → SVG (vector) + PNG (raster)
- **Slide content** → PPTX
- **Marketing copy block** → social-media-sized image (PNG) at the right aspect ratio + raw text file
- **Dashboard / KPI panel** → SVG with editable layers

### Step 2 — Pick the target format

Ask if not stated, otherwise infer from intent:

| Target use | Format | Notes |
|---|---|---|
| Figma import | SVG (vector) + PNG (raster preview) | SVG preserves editable layers |
| Canva import | PNG (transparent, 2x) + raw text in TXT | Canva imports both |
| Adobe (Illustrator/Photoshop) | SVG, AI, or layered PSD-like | Default to SVG when AI/PSD not available |
| PowerPoint | PPTX direct | Use the `pptx` skill for layout |
| Word / Google Docs | DOCX | Use the `docx` skill |
| Web publishing | PNG (2x retina) + WebP | Optimize file size |
| LinkedIn post | PNG 1200×627 or 1080×1080 | |
| Instagram post | PNG 1080×1080 (feed) or 1080×1920 (story) | |
| LinkedIn / X carousel | PDF, multi-page; each page 1080×1350 | |
| Email | PNG inline + alt text | < 1MB |

### Step 3 — Export with brand variants

For each export, produce at least two variants:

1. **Brand-aware** — applies the user's brand system (colors, typography, logo). For ACTCENTER: dark mode (Carbon `#0E1117` background, Signal Green `#00FF88` accent, Space Grotesk + Inter + JetBrains Mono).
2. **Neutral** — no brand styling, designer-ready (transparent background, neutral palette, vector layers).

Optionally a third **light-mode** variant when the destination requires it (e.g., printed PDF, light email template).

## How to run

```
1. Confirm: source asset + target tool + target format + brand variant or not.
2. Load the appropriate skill if the format demands it:
   - .pptx → load pptx skill
   - .docx → load docx skill
   - .xlsx → load xlsx skill
   - .pdf → load pdf skill
3. For raster (PNG): generate at 2x density, transparent background by default.
4. For vector (SVG): clean groups, named layers, no rasterized text.
5. Save to /skills-downloads/ or the user's chosen folder.
6. Present each export with a computer:// link.
```

## Naming convention

```
{asset-slug}_{variant}_{format}_{size}.{ext}

examples:
  q1-revenue-chart_brand_png_2x.png
  q1-revenue-chart_neutral_svg.svg
  q1-revenue-chart_brand_linkedin-1200x627.png
  weekly-report_brand_pptx.pptx
```

## Brand presets — ACTCENTER

When the user is exporting for ACTCENTER, default to:

- **Background:** Carbon `#0E1117` (dark) or Off-White `#F0F6FC` (light)
- **Accent:** Signal Green `#00FF88` for OK / positive
- **Critical:** Alert Red `#FF3838`
- **Warning:** Bright Amber `#FFB800`
- **Secondary bg / cards:** Petroleum `#0E3A4A`, Steel Gray `#1A1F2E`
- **Display:** Space Grotesk Bold
- **Body:** Inter Regular
- **Data / monospace:** JetBrains Mono
- **No stock photography of people.** If illustration is needed, use abstract geometry, dashboard mockups, or terminal-style imagery.
- **No emojis in business comms.**

If the user is exporting for another brand, ask for the brand system before defaulting.

## Output template

For each export job, ship this back:

```
## Export Manifest — [asset name] — [YYYY-MM-DD]

### Source
- Type: [report / chart / table / diagram / copy block]
- Origin: [conversation block / artifact / file path]

### Target
- Destination tool: [Figma / Canva / Adobe / PPT / Web / LinkedIn / IG / Email]
- Format(s): [SVG / PNG / PDF / PPTX / DOCX]
- Variants delivered: [brand / neutral / light-mode]

### Files
- [filename_1] — [computer:// link]
- [filename_2] — [computer:// link]

### Notes for the designer
- Vector layers named: [yes/no]
- Fonts: [embedded / outlined / live text — and which fonts]
- Color profile: [sRGB / display-P3]
- Cropping safe-zones: [if social-media format]
```

## Voice

Operational. Show the manifest, not the artist's note. The user wants files + links + a one-line note per file, not a creative rationale.

## Boundaries

- Don't redesign the asset on export. Export it; let the designer iterate.
- Don't compress aggressively below the format's quality floor (e.g., don't push PNG below 80% if the destination is print or retina).
- Don't strip the brand layer silently. If the user asks for "neutral," confirm before stripping.
- Don't email or upload the files anywhere. Save locally and present links — the user owns the distribution.