Testing & QA

Avoid common debugging traps by running the decisive experiment first

Chain evidence-linked whys to find the systemic root cause of a fault

Narrow a defect with IS/IS-NOT analysis and weigh options by must and want

Rank falsifiable hypotheses and run the cheapest test that tells them apart

Root-cause wrong results by reproducing, minimizing and instrumenting before fixing

Compare two implementations such as R and Python ports to confirm they agree

Prove a checker catches seeded defects and measure its recall and false positives

Run browser-based smoke, regression and user-flow tests against live websites

Run and debug a local web app with a managed browser and a combined event timeline

Build and report an accessibility coverage matrix across criteria and surfaces

Debug by forming testable hypotheses and confirming or rejecting each one

Explore several root-cause hypotheses in parallel across system areas

Trace a bug or crash to its true root cause instead of treating symptoms

Apply AAA, Given-When-Then, test builders, fixtures and doubles to test suites

Find flaky tests, over-mocking and brittle assertions and fix them

Narrow down exactly where a bug starts with tracing and isolation steps

Verify a change against its spec in the real app and review it with a fresh model

Find and fix a bug's root cause with a reproduce, test and verify loop

Diagnose brittle tests, mock abuse and coverage gaps in an existing test suite

Debug network, streaming and TLS failures with falsifiable layered experiments

Audit rendered web and app UI through browser journeys and Playwright sweeps

Set up QA processes with test strategies, bug severity tracking and handoff docs

Test TypeScript snippets before saving them as reusable skills

Write Spring Boot tests with JUnit 5, Mockito, MockMvc and Testcontainers

Generate pytest tests for Typer CLI commands with CliRunner fixtures and edge cases

Plan and write unit, integration and E2E tests, and improve coverage

Diagnose a failure by testing rival hypotheses in parallel and ranking evidence

Check whether a past engineering decision still holds by comparing baseline and results

Debug test failures and unexpected behavior systematically before proposing fixes

Write failing tests first and drive each change with strict test-driven development

Verify work actually passes before claiming it is fixed, complete or ready to merge

Capture before and after screenshots of pages or elements for PR visual comparison

Test UI behavior hands-on with screen recordings and assertion annotations as proof

Prune redundant tests and decide which suite owns each contract

Compare screenshots against intended designs to catch visual regressions

Open screenshots and renders together in one macOS Preview window

Write tests that pin real behavior instead of implementation details

Write pytest suites that cover edge cases and guard against regressions

Set up an end-to-end test suite with auth helpers, video traces and layered assertions

Set up a repo so engineering work is proven to work before it ships

Review generated tests against universal testing rules before they are committed

Make a Simulink model compatible with Design Verifier by fixing unsupported constructs

Inject stuck, noise and offset faults into Simulink signals to test robustness

Find and fix division by zero, overflow and dead logic in Simulink designs

Explain why Simulink coverage objectives are missing from an existing result

Write Gherkin-based and persistent Simulink tests from requirements or specs

Run systematic exploratory QA on web apps, capture evidence and report bugs

Find the root cause of a bug or test failure before proposing any fix

Write failing tests first and follow the red, green, refactor cycle