agentic-browser-testing
Test changing app flows with a browser agent.
What it does
Creates bounded browser-agent tests with independent outcome checks, evidence, and a path to scripted regression tests.
When to use it
Use for fast-changing interfaces, hard-to-locate flows, and exploratory smoke tests; keep stable critical paths scripted.
How to use it
Give it an application flow to test; it creates a bounded browser-agent run with independent checks, a trace, and an evidence report.
What you provide
- the user journey to test
- An existing project
Uses
Good to know
- starts helper agents, number unstated
What you need · 6
The snapshot-first browser workflow uses Playwright MCP to navigate, inspect the accessibility tree, click, type, and wait.
Requires a project-owned host adapter to run the browser agent and enforce budgets; no model runner is bundled.
Declared a universal dependency supplying stack, environment, seed/reset tooling, and model-access context needed by discovery.
Access · 1
This skill
@playwright/test
Installs Playwright Test