skills.rest
bacoco
bacocoCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

sg-visual-run

Execute Playwright/agent-browser YAML visual tests and detect UI regressions.

7|2|Updated Mar 24, 2026
Search Tags:#playwright#ui-testing#visual-testing#test-automation#regression-testing#agent-browser#llm-assertions
One-click install
npx skills add https://github.com/bacoco/ShipGuard --skill sg-visual-run
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sg-visual-run
Source: https://github.com/bacoco/ShipGuard/tree/main/plugins/shipguard/skills/sg-visual-run
Command: npx skills add https://github.com/bacoco/ShipGuard --skill sg-visual-run

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Automates execution of visual test manifests to detect UI regressions, broken layouts, and visual errors so teams do not rely on manual screenshot inspection or ad-hoc checks.

Core Features & Use Cases

  • Manifest-driven execution: Run YAML visual tests with agent-browser to verify pages, flows, and uploaded files across environments.
  • Hybrid assertions: Combine mechanical Playwright-style steps with LLM-powered llm-checks and mandatory screenshot validation to catch semantic and visual issues.
  • Robust session and recovery: Session-expiry detection, sequential single-session execution, browser crash recovery, regression bookkeeping, and human-readable reports for CI and interactive runs.
  • Use case: After modifying a header or sidebar, run targeted diffs or regressions to confirm no layout breakages, capture failing screenshots, and update the regression history.

Quick Start

Run the visual test suite for changed pages by invoking /sg-visual-run with a natural language description such as I changed the sidebar, check it still works

Frequently Asked Questions about sg-visual-run

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I detect UI regressions after modifying a layout component?▼

Detect UI regressions by running visual end-to-end tests with agent-browser YAML manifests that capture targeted diffs and confirm no layout breakages occurred after your modifications.

Can I use LLM assertions for semantic visual checks in Playwright?▼

Yes, you can use LLM-assisted assertions combined with mechanical Playwright-style steps and mandatory screenshot validation to catch both semantic and visual issues during test automation.

How do I run visual regression tests in CI environments?▼

Run visual regression tests in CI by executing sequential agent-browser sessions with session-expiry detection, crash recovery, and regression bookkeeping to generate human-readable reports for automated pipelines.

What is the best way to recover from browser crashes during automated visual testing?▼

The best way to recover from browser crashes during automated visual testing is using sequential single-session execution with built-in crash recovery that automatically resumes interrupted visual test runs.

Does agent-browser support sequential sessions for regression testing?▼

Yes, agent-browser supports sequential single-session execution for regression testing, featuring mandatory screenshot reading, session-expiry detection, and regression history bookkeeping in visual-tests/_regressions.yaml.

Why do my visual tests fail to catch broken layouts after header changes?▼

Visual tests fail to catch broken layouts when lacking mandatory screenshot validation; applying hybrid LLM-powered checks alongside mechanical Playwright steps surfaces both semantic and visual layout errors.

Related Skills

Explore More Software Engineering Skills→
bacocobacoco
7

sg-record

Record browser interactions and export ShipGuard YAML test manifests.

Community
Advanced
bacocobacoco
7

sg-visual-review

Generate an interactive HTML review page for visual test results.

Community
Advanced
gandazgulgandazgul
14

playwright-cli

Automate browser interactions and generate Playwright test scripts.

Community
Advanced
tyronerosstyroneross
1

validate

Audit UI components and interactions against accessibility standards via CDP.

Community
Advanced