c-screen — Claude Code skill for screenshots and OCR
Capture screen images, extract on-screen text with OCR, or take webcam frames for visual inspection. Useful when an agent needs to examine what’s visible on a display or camera.
npx skills add daxaur/openpaw --skill c-screen -a claude-codeWhat c-screen does
The skill uses peekaboo to capture a full screen, a named app window, or a selected region, and to extract visible text with OCR. It uses camsnap for webcam images, but only when the user explicitly asks for one. After capture, the agent reads the image file to inspect its contents.
peekaboo needs macOS screen recording permission; camsnap needs camera permission. OCR results depend on font size and screen resolution, so small text can be harder to extract. To install for Claude Code, run npx skills add daxaur/openpaw --skill c-screen -a claude-code; the skill file is placed at ~/.claude/skills/c-screen/SKILL.md.
What it helps with
- capture the full screen for visual inspection
- extract visible text from a screen with OCR
- save an image of a specific app window
- capture a frame from a selected webcam
Example prompts for the agent
- Capture my screen and describe what is visible.
- Extract the text currently shown on screen.
- Take a webcam image and inspect its contents.
How to install c-screen
1Skills CLI (Vercel)
npx skills add daxaur/openpaw --skill c-screen -a claude-code2Manually: put the files in place
~/.claude/skills/c-screen/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/c-screen/
1Skills CLI (Vercel)
npx skills add daxaur/openpaw --skill c-screen -a codex2Manually: put the files in place
.agents/skills/c-screen/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add daxaur/openpaw --skill c-screen -a cursor2Manually: put the files in place
.cursor/skills/c-screen/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install daxaur/openpaw c-screen2Manually: put the files in place
.github/skills/c-screen/SKILL.md1Skills CLI (Vercel)
npx skills add daxaur/openpaw --skill c-screen -a gemini-cli2Platform CLI
gemini skills install https://github.com/daxaur/openpaw --path skills/c-screenManually: put the files in place
.devin/skills/c-screen/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add daxaur/openpaw --skill c-screen -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about c-screen
How do I install c-screen in Claude Code?
Run npx skills add daxaur/openpaw --skill c-screen -a claude-code. The skill file is available at ~/.claude/skills/c-screen/SKILL.md.
What does c-screen do?
It uses peekaboo to capture screen images and run OCR, and camsnap to take webcam images. The agent can then read and analyze the captured image.
What permissions does c-screen need?
peekaboo requires macOS screen recording permission, while camsnap requires permission to use the camera.
Source text
SKILL.md · MIT
--- name: c-screen description: Capture screenshots and extract text via OCR using `peekaboo`, and capture webcam images using `camsnap`. Enables visual analysis of screen content and camera input. tags: [screenshot, ocr, webcam, peekaboo, camsnap, vision] --- ## What This Skill Does Enables Claude to take screenshots, extract text from the screen via OCR, and capture webcam images for visual analysis using `peekaboo` and `camsnap`. ## Available CLI Tools ### `peekaboo` — Screenshots & OCR ```bash # Take a full screenshot and save to file peekaboo screenshot --output ~/Desktop/screen.png # Take screenshot and extract all text via OCR peekaboo ocr # OCR a specific region (x, y, width, height) peekaboo ocr --region 0,0,1280,720 # Screenshot a specific window by app name peekaboo screenshot --app "Safari" --output window.png # Screenshot and pipe to stdout for immediate analysis peekaboo screenshot --stdout ``` ### `camsnap` — Webcam Capture ```bash # Capture a single webcam frame camsnap --output ~/Desktop/photo.jpg # Capture with a specific camera device camsnap --device 0 --output shot.jpg # Capture and print file path camsnap --output /tmp/cam.jpg && echo "Saved" ``` ## Usage Guidelines - Use `peekaboo ocr` when the user wants text extracted from the screen - Use `camsnap` only when the user explicitly wants a webcam image - After capturing, read the image file to analyze its contents visually ## Notes - macOS screen recording permission required for `peekaboo` - Camera permission required for `camsnap` - OCR accuracy depends on font size and screen resolution
© daxaur · MIT · daxaur/openpaw