Skip to content
Development
Skill

/visual-validation

Autonomous screenshot verification — captures, compares, and AI-evaluates UI changes for visual regressions

From plugin
maggy
71073 skills1 agent25 commands1 hook

Claude Code

$ npx -y skills add alinaqi/maggy --skill visual-validation --agent claude-code

SKILL.md

visual-validation.SKILL.md
name: visual-validation
description: Autonomous screenshot verification — captures, compares, and AI-evaluates UI changes for visual regressions
when-to-use: When a UI or static-asset change needs visual verification before shipping
user-invocable: false
effort: medium

Visual Validation — Autonomous Screenshot Verification

Philosophy

Every UI change should be visually verified before it ships. Peekaboo captures pixel-accurate screenshots. The system compares before/after and flags visual regressions. No manual "looks good to me" — the machine verifies what the machine built.

Autonomous Flow

static/* files modified (detected by auto-review-hook or E2E testkit)
    ↓
peekaboo image --mode screen → ~/.maggy/visual-verify/after-{ts}.png
    ↓
Compare with latest baseline (previous after-*.png)
    ↓
AI evaluation: Gemini Flash analyzes screenshot for visual regressions
    ↓
Report: CLEAN or REGRESSION_DETECTED with specific issues

Quick Commands

# Install
brew install steipete/tap/peekaboo

# Capture full system screenshot
peekaboo image --mode screen --retina --path /tmp/screen.png

# Capture specific window
peekaboo image --mode window --app "Google Chrome" --path /tmp/window.png

# Verify Maggy dashboard after code change
peekaboo image --mode window --app "Google Chrome" --path ~/.maggy/visual-verify/after-$(date -u +%Y%m%d-%H%M%S).png

# Compare two screenshots
ls -la ~/.maggy/visual-verify/after-*.png  # review before/after pairs

Autonomous Verification Flow (Claude Code)

# After making any static/ file change, Claude should:
1. git add maggy/maggy/static/ && git diff --stat HEAD  # check what changed
2. Restart Maggy: lsof -ti :8080 | xargs kill && python3 -m maggy.main &
3. Wait for health: curl -s http://localhost:8080/api/health
4. Capture screenshot: peekaboo image --mode screen --path ~/.maggy/visual-verify/after-$(date -u +%Y%m%d-%H%M%S).png
5. Compare with latest baseline in ~/.maggy/visual-verify/
6. AI evaluation: ~/bin/gemini-api --flash "Analyze this screenshot for visual regressions"
7. Report findings

Maggy Dashboard Verification

# One-shot verification
curl -s http://localhost:8080/api/health && \
peekaboo image --mode screen --path ~/.maggy/visual-verify/verify.png && \
echo "✓ Dashboard verified — screenshot at ~/.maggy/visual-verify/verify.png"

Integration Points

Auto-Review Hook (already wired)

  • Detects `static/` changes in git diff
  • Auto-captures Peekaboo screenshot
  • Saves before/after pairs in `~/.maggy/visual-verify/`

E2E Testkit Plugin

  • Visual regression testing as part of `POST /api/testkit/run`
  • Compares screenshots against baselines
  • Flags pixel differences above threshold

Build-in-Public Plugin

  • Peekaboo screenshots replace AI-generated images
  • Real product captures for LinkedIn/X posts

When to Use

| Scenario | Command | |----------|---------| | After static/ change | `peekaboo image --mode screen --path ~/.maggy/visual-verify/after.png` | | Before committing UI change | Run verification flow above | | Reviewing dashboard | Screenshot + AI analysis | | Build-in-public post | Auto-captured by plugin | | Cross-browser check | `peekaboo image --mode window --app "Safari"` |

Read more
@alinaqi
Ships withmaggy

Turn Claude Code into a self-reviewing, test-enforced engineering system that remembers context across sessions — then route work across 13 models from a single dashboard.

71073 skills1 agent
Get the whole plugin

Stars on GitHub

Momentum

stars in total
710
Star history of alinaqi/maggy on GitHub, a chart by star-history.com

Other skills on maggy.

aeo-optimization

aeo-optimization

AI Engine Optimization - semantic triples, page templates, content clusters for AI citations

@alinaqi@alinaqiInstall Skill
agent-teams

agent-teams

Claude Code Agent Teams - default team-based development with strict TDD pipeline enforcement

@alinaqi@alinaqiInstall Skill
ai-models

ai-models

Latest AI models reference - Claude, OpenAI, Gemini, Eleven Labs, Replicate

@alinaqi@alinaqiInstall Skill
android-kotlin

android-kotlin

Android Kotlin development with Coroutines, Jetpack Compose, Hilt, and MockK testing

@alinaqi@alinaqiInstall Skill

More skills from other plugins.

brainstorming
Auto-invoked

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or…

@obra@obraInstall Skill
ask-matt

ask-matt

Ask which skill or flow fits your situation. A router over the skills in this repo.

skills2w309k
@mattpocock@mattpocockInstall Skill
accessibility
Auto-invoked

accessibility

Design, implement, and audit accessible UI to WCAG 2.2 Level AA across Web, iOS, and Android — semantic ARIA roles and labels,…

ecc2w293k
@affaan-m@affaan-mInstall Skill
compiler-commit

compiler-commit

Use when you want to verify compiler changes and commit with the correct convention. Runs tests, lint, and format, then commits…

react5d252k
@facebook@facebookInstall Skill

Questions about visual-validation.

Which coding agents can use visual-validation?
visual-validation installs in Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode and Windsurf with the skills CLI, using each agent's flag shown on this page. Flowy's auto-invocation, which fires a skill when your prompt matches, runs in Claude Code only.