playwright-autopilot

Use this agent to investigate a single failing E2E test. Performs read-only deep-dive using DOM snapshots, network traces, console output, and screenshots to produce a structured diagnosis. Never modifies code. Examples: <example>Context: A test failure needs root cause analysis. user: 'Investigate why the checkout test at tests/checkout.spec.ts:42 is failing' assistant: 'I will use the bug-investigator agent to analyze the failure evidence and produce a diagnosis.' <commentary>Deep investigation of a single failure, producing a diagnosis report for the fixer.</commentary></example><example>Context: Triage identified a new failure that needs understanding. user: 'This test started failing after the deploy, figure out why' assistant: 'I will use the bug-investigator to trace the root cause through DOM state, network, and console logs.' <commentary>Read-only investigation to understand what changed.</commentary></example>

Use this agent to apply a code fix to a failing E2E test based on an investigation diagnosis. Takes a structured diagnosis and applies the minimal change to make the test pass. Runs in a worktree for parallel safety. Examples: <example>Context: An investigator diagnosed a missing test step. user: 'Fix checkout.spec.ts:42 — needs selectShippingMethod call after line 42' assistant: 'I will use the test-fixer agent to apply the fix and verify the test passes.' <commentary>Targeted code fix based on a diagnosis.</commentary></example><example>Context: A test has a stale selector after UI changes. user: 'The login test selector needs updating from role button Submit to role button Sign In' assistant: 'I will use the test-fixer agent to update the selector and verify.' <commentary>Simple code fix with clear instructions.</commentary></example>

Use this agent to write new Playwright E2E tests from scratch by interactively exploring the application. Navigates pages, discovers user flows, and creates production-grade tests with POM/business-layer architecture. Runs in a worktree for parallel safety. Examples: <example>Context: User needs tests for a new feature. user: 'Write e2e tests for the new checkout flow' assistant: 'I will use the test-writer agent to explore the checkout flow and create comprehensive tests.' <commentary>New feature needs end-to-end test coverage from scratch.</commentary></example><example>Context: Coverage gaps identified by triage. user: 'We have no tests for the settings page' assistant: 'I will use the test-writer agent to explore the settings page and write tests for it.' <commentary>Missing test coverage for an existing feature.</commentary></example>

Use this agent to triage the full E2E test suite. Runs all tests, classifies every failure, manages project knowledge (error patterns, flows, triage history), creates tasks, and dispatches investigator/fixer agents. Examples: <example>Context: User wants a full suite health check. user: 'Triage the e2e suite' assistant: 'I will use the triage-lead agent to run all tests, classify failures, and coordinate fixes.' <commentary>Full suite triage with failure classification and parallel fixer coordination.</commentary></example><example>Context: CI is red and needs investigation. user: 'CI tests are failing, figure out what is broken' assistant: 'I will use the triage-lead agent to run the suite, identify root causes, and dispatch fixers.' <commentary>Suite-level investigation requires the triage coordinator.</commentary></example>

Systematically investigate and fix failing Playwright E2E tests using captured action data, screenshots, DOM snapshots, network requests, and console output.

Run all E2E tests, categorize failures (known issue, app bug, test update, flaky, new), cross-reference Jira, and generate a management-ready report.

Write new Playwright E2E tests following project conventions, with POM/business-layer architecture, network-aware stability patterns, and quality validation.