Testing & QA

Write test scenarios with starting conditions, steps and expected outcomes from stories

Run a diagnosis loop on failing or slow behavior to find the real cause

Browser QA the branch diff, fix small breakages and write a dogfood report

Run browser tests for the pages touched by the current branch or PR

Trace errors and bugs to their root cause before touching any code

Group new Sentry errors, dedupe against GitHub issues and draft tickets for the top ones

Quarantine flaky tests with skip reasons and a tracking issue based on CI history

Check out each new pull request, run the full suite and post pass or fail results

Harden tests with property-based and mutation testing, scoring and DRY scans

Find systemic root causes with Five Whys, fishbone and blameless postmortems

Add an inspector panel to a json-render app to debug specs, state and actions

Open the Windmill dev page to visually check a flow, script or app as you build it

Test REST and GraphQL APIs with curl, auth flows and debugging steps

Search GitHub, Stack Exchange and developer docs for error fixes and workarounds

Diagnose Sentry errors by mapping stack frames to your local source code

Write Playwright end-to-end tests with proper waits and API checks

Validate UI changes in a branch with Playwright and add missing end-to-end tests

Find the root cause of a failing test or runtime bug before trying any fix

Run real verification commands and show evidence before claiming a task is done

Write Playwright end-to-end tests with page objects and reliable selectors

Write pytest tests with fixtures, mocks, parametrize and markers

Write Vitest unit tests for React components, hooks and utilities with RTL

Automate browser steps, record Playwright code and inspect selectors from the CLI

Run compile and type-check commands and report every failure clearly

Drive and profile an interactive CLI or TUI to catch hangs and prompt bugs

Drive a web, IDE or Electron UI through a local browser harness for QA

Run Playwright smoke tests, debug failures and verify the fixes

Draw app states as Mermaid statecharts from TLA+ model checking results

Verify a claim against baseline and treatment evidence before trusting it

Review code changes for test coverage gaps, edge cases and regressions

Design chaos experiments, failure injection and game days for distributed systems

Trace errors and stack traces to a root cause with a hypothesis-driven method

Write Playwright end-to-end tests with page objects, CI setup and flaky test fixes

Generate unit, integration and E2E tests with mocks, coverage analysis and test plans

Generate a project-local skill that launches your app and bug bashes each feature

Write, run and debug end-to-end tests for web and mobile apps with an agentic runner

Port Playwright, Cypress, Selenium or Detox suites to e2e one slice at a time

Validate or debug a workflow in the browser with reproducible evidence

Reproduce, isolate and fix bugs in a structured debugging session

Design test strategies, test plans and coverage targets for your project

Record captioned demos for each critical user journey through its real entry point

Find a bug's root cause, make a focused fix and prove it with targeted tests

Verify a reported bug's path, cause and history before choosing a fix

Audit a recovered repro patch, its assertions and baseline before reviewing it

Prove user-facing behavior with recorded evidence by driving the app like a user

Generate tests for a finished project phase from its UAT criteria and code

Audit all outstanding UAT and verification items across project phases

Debug systematically, keeping state across context resets and compaction

Investigate failed project workflows after the fact and find what went wrong