agentleFS
seancabalseseancabalse/agentic-playbook/skills/capture-screenshots/SKILL.md

capture-screenshots

Skill0 starsChanged 15 days agoMITView on GitHub

Capture screenshots of each state, size and setting as evidence for a PR. Use for each UI change.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Files
SKILL.md

Capture screenshots

Playbook files named here: when the project has its own copy at .playbook/local/<same path>, use that copy instead.

Request: $ARGUMENTS

Drivers and their traps: ${CLAUDE_PLUGIN_ROOT}/principles/08-testing-qa.md → "Drivers". Ready-made scripts: ${CLAUDE_PLUGIN_ROOT}/templates/testing/.

1. Decide the matrix (keep it to what the change touches)

DefaultAdd when it matters
Statesthe populated defaultloading, error, offline, empty
Sizes375 px mobile and 1280 px desktop320 px (reflow), 768 px tablet, 200% zoom
Settingslight, normal motiondark (colorScheme), reduced motion

2. Capture

  • Web, with Playwright: copy templates/testing/playwright.config.ts and screenshots.spec.ts (if the project has no equivalent), list the routes and states, and run npx playwright test screenshots.spec.ts. To explore before scripting, use the agent's browser tools (the built-in browser, Claude in Chrome or the Playwright MCP server), then put the steps in the spec so they're repeatable.
  • Zoom: 200% browser zoom on a 1280 px window is a 640 CSS px viewport (add deviceScaleFactor: 2 for sharp text), and 400% is 320 px. Capture those viewports.
  • Name the files <screen>-<state>-<variant>.png (settings-export-progress-zoom200.png) and put them in screenshots/<ticket>/, which is gitignored.

3. Check every image

Open each one and look:

  • Is it the right screen and state?
  • Is any text clipped or overlapping?
  • Is a dev overlay or a loading spinner covering it?
  • Is a system dialog in the way?

A blank or half-drawn image usually means the page hadn't finished loading, a request was blocked, or the capture ran mid-animation: fix the cause and recapture. Never ship an image you haven't looked at.

4. Hand over

  • List the files with one line each (what it shows) in your handoff.
  • ship-pr publishes them: templates/testing/push-screenshots.sh <PR> <files…> pushes to a screenshots branch, never the default branch.
  • Reset every setting you changed (zoom, colour scheme, reduced motion), and stop what you started.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.