Automated visual QA testing using Playwright — navigate web apps like a real user, capture screenshots, find bugs, and fix them.
SkillScore73Maintained
Automatic
◆QA & Test Engineering pack
Last verified 8/20/2026
What it changes
Passed 14 of 19 test cases with the skill, and 5 of 19 without it.
Without skill27%
With skill73%
Checks it passed that the model alone missed
Names qa_cleanup as the mandatory final step.
Specifies using qa_navigate to open specific routes.
Specifies calling qa_check to run an individual check.
and 9 more in the full report
10 fixed6 passed either way6 still failing0 broke
Pass-rate lift
+45 pp
Tokens
+11%
Agent turns
0%
Response time
-63%
Model gemini-3.6-flashMeasured 8/20/2026v1 · latest is v216 of 22 cases passed with skill · 19 comparable
Tokens, turns and time describe resource use, not a dollar price.
This benchmark measured an earlier version of the skill.
See it in action
One of 10 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 10
Prompt
An automated browser testing sequence and issue resolution session on http://localhost:3000 has completed. What tool call must always be invoked as the final step of the session?
What was checked, without and with the skill
Checked
Without skill
With skill
Names qa_cleanup as the mandatory final step.
Missed
Passed
case-13 · gemini-3.6-flash · 8/20/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md109 linessingle file
Playwright QA
Purpose
Automated visual QA testing for web applications. Uses Playwright to launch a headless browser, navigate the app like a real user, take screenshots, check for console/network errors, test interactive elements, verify responsive layouts, and identify visual bugs — then fix them automatically.
Routing
Use when: Any web app task that mentions testing, catching bugs, verifying it works, shipping, or quality — even without explicitly mentioning Playwright or visual QA. Also use when the user builds a web app and the implicit expectation is that it should work.
Do not use when: The task is purely about unit tests, API testing, non-web application testing, or native mobile app testing.
Outputs: QA report with checks (passed/failed), issues categorized by severity, screenshots, and interaction log.
Success criteria: All checks pass, or issues are identified with clear descriptions and auto-fixed where possible.
Trigger Examples
Positive
Build a todo app in React, test it to catch any bugs before shipping
Build a dashboard and make sure it works
Create a landing page and verify everything looks right