opencode-skill-image-to-code
v1.0.0
Published
OpenCode skill: 100% programmatic image-to-code converter via OCR + pixel analysis (no AI vision needed)
Maintainers
Readme
opencode-skill-image-to-code
100% programmatic image-to-code converter for OpenCode. No AI vision needed — extracts text (OCR), colors, layout sections, and UI components using code only.
Install
npm install -g opencode-skill-image-to-codeUsage
# Analyze image file
& "$env:USERPROFILE\.agents\skills\image-to-code\analyze-image.ps1" -ImagePath "screenshot.png"
# Analyze from clipboard
& "$env:USERPROFILE\.agents\skills\image-to-code\analyze-image.ps1" -ClipboardScripts
| Script | Purpose |
|--------|---------|
| analyze-image.ps1 | Full pipeline — OCR + colors + layout |
| ocr.ps1 | Text extraction with optional bounding boxes |
| extract-colors.ps1 | Color palette + semantic color detection |
| detect-layout.ps1 | Layout sections + component identification |
