Agent skill
extracting-stitch-mockups
Extracts generated mockup images from Google Stitch project pages. Use when user provides a Stitch project URL (stitch.withgoogle.com/projects/...), mentions extracting/downloading Stitch mockups, saving Stitch designs, or wants to archive generated design assets. Requires authenticated Chrome browser profile.
Stars
163
Forks
31
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/extracting-stitch-mockups
SKILL.md
Extract Stitch Mockups
Quick Start
- Get project URL - User provides
https://stitch.withgoogle.com/projects/{id} - Resolve feature directory - Determine where to save using fallback chain
- Run extraction script - Execute
scripts/extract_images.pywith Playwright - Save to exports - Images saved to
design-intent/google-stitch/{feature}/exports/
Feature Directory Resolution
Determine target feature directory using this fallback chain:
- User specifies - Optional
--featureargument provided - Auto-detect - Match Stitch project title to existing
design-intent/google-stitch/{feature}/directories - Prompt user - List existing directories and ask user to select or create new
Auto-Detection Logic
- Extract project title from Stitch page (e.g., "Eco-Travel Home Screen")
- Normalize to feature format: lowercase, hyphens, strip special chars
- Search for matching directory in
design-intent/google-stitch/ - If multiple partial matches, prompt user to select
Extraction Process
Prerequisites
- Chrome browser with active Google session
- uv installed (https://github.com/astral-sh/uv)
- Playwright browsers:
uv run playwright install chromium
Script Execution
bash
# Basic usage
uv run scripts/extract_images.py "https://stitch.withgoogle.com/projects/123"
# With explicit feature directory
uv run scripts/extract_images.py "https://stitch.withgoogle.com/projects/123" --feature dashboard
# Or run directly (after chmod +x)
./scripts/extract_images.py "https://stitch.withgoogle.com/projects/123"
Image Filtering
- Source:
lh3.googleusercontent.com/aida/...URLs only - Size filter: Images with dimensions >= 400px (mockups, not UI elements)
- Excludes: avatars, icons, UI chrome
Generation Check
Script checks for "Generating..." status on page:
- If detected: Exit with message to retry after generation completes
- If complete: Proceed with extraction
Output Structure
Images saved to existing feature's exports/ directory:
design-intent/google-stitch/{feature}/exports/
├── mockup-1.png
├── mockup-2.png
└── mockup-N.png
File Naming
- Sequential numbering:
mockup-{index}.png - Index starts at 1
- Preserves original image format (typically PNG)
Report
After extraction, display summary:
Extracted {N} mockups from Stitch project
Project: {project-title}
URL: {project-url}
Saved to: design-intent/google-stitch/{feature}/exports/
- mockup-1.png (400x800)
- mockup-2.png (400x800)
- ...
Feature directory:
design-intent/google-stitch/{feature}/
├── prompt-v{N}.md
├── exports/ <- Mockups saved here
│ ├── mockup-1.png
│ └── mockup-2.png
└── wireframes/
Common Issues
- Not authenticated - Open Chrome, sign into Google, then retry
- Still generating - Wait for Stitch to complete, then retry
- No feature directory - Run authoring-stitch-prompts first, or specify
--feature
See TROUBLESHOOTING.md for detailed solutions.
Reference Files
- WORKFLOW.md - Detailed browser automation steps
- EXAMPLES.md - Sample extractions
- TROUBLESHOOTING.md - Error handling
Didn't find tool you were looking for?