✅ Testing & QA

Verifies UI changes in a real browser before declaring them done — drive the changed flow, check the console, test the mobile viewport, and report what was actually seen. Use whenever a change touches anything a user sees or clicks, and before any "fixed" or "done" claim about frontend work.

Traces every button, toggle and form submit in an area through each state read and write in order, after first mapping what every store action sets and silently resets, to find handlers that work call by call but cancel each other out, end in a state the label doesn't promise, or race. Use when a button "does nothing" with no error, when users report broken controls that debugging didn't find, or after changing a shared state store or refactoring shared state.

Derive a 5-15 check test plan from a diff, run it in a real browser, and report issues with evidence.

Hands work to the human only after verifying everything the agent can itself, then gives a short, risk-ordered list of the checks only a person can do (real devices, their accounts and inboxes, live payments, production, taste and business judgment), each with exact steps, the expected result, and why it needs a human. Use after any major code push, finished feature, opened PR, merge, or deploy, and whenever about to ask the user to test, try, check, confirm, or "let me know if it works".

promptfoo test cases, calibrated graders, and a GitHub Actions gate so prompt or model changes cannot regress.

Turns "it seems to work" into "it passes my tests". It writes 8 to 10 test cases for anything built in this chat (a tool, a prompt, a voice guide, a brief, a report), runs them honestly, fixes what fails and reruns until every test passes, then saves the tests so you can rerun them after any change. Use when someone wants to check AI work before trusting it, or wants to learn what an eval is by doing one.

Vitest 5 + Playwright 1.64 + Testcontainers: a fast unit core, real-database integration tests, few E2E flows.

Gets verification done in apps that sit behind sign-in — via the app's own dev bypass, a temporary identity, or the user's already-signed-in browser. Apply when a login wall blocks opening, testing, or screenshotting an app, or when about to skip browser verification because the app requires authentication.

Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.