Invoke the playwright-enterprise-tester skill to write, run and fix Playwright E2E tests. Supports modes smoke/critical-path/ajax-heavy/auth-protected/visual-regression/perf-budget/legacy-mpa/release-gate plus 5 phase 2 modes (dynamic-scraping/a11y-scan/mobile-perf/cross-browser/popup-checkout). Scope selectable by file, folder, glob, tag, or mode. Multi-stack autodetect. Defensive-minimal defaults, 17 opt-in phase 2 features via .playwright-tester.json.

/playwright-tester

Invoke the playwright-enterprise-tester agent to work on your project's Playwright E2E tests.

This slash command is a thin wrapper that:

Parses user arguments
Resolves the runner mode (local or ci)
Loads the playwright-enterprise-tester skill
Delegates execution to the dedicated agent

Arguments

Pass arguments as key=value pairs separated by spaces. None are required.

Argument	Example	Description
`mode=`	`mode=critical-path`	Operating mode (default: `smoke`)
`files=`	`files=tests/e2e/cart.spec.ts,tests/e2e/checkout.spec.ts`	CSV of spec file paths
`folders=`	`folders=tests/e2e/critical/,tests/e2e/smoke/`	CSV of folders
`grep=`	`grep=checkout`	Playwright `--grep` pattern
`tags=`	`tags=@critical,@smoke`	Tag list (mapped to `--grep`)
`brand=`	`brand=mybrand`	Override multi-tenant brand
`country=`	`country=us`	Override multi-tenant country
`lang=`	`lang=en`	Override multi-tenant language
`update-snapshots=`	`update-snapshots=true`	Visual regression only, forbidden in CI
`fix-app-code=`	`fix-app-code=true`	Requires double-confirm governance
`dry-run=`	`dry-run=true`	Plan without executing
`runner=`	`runner=local` or `runner=ci`	Override autodetect

Scope precedence

files > folders > grep > tags > scope inferred from mode > single smoke run.

If no args are passed and the scope is ambiguous, the agent asks for confirmation before running. The full suite is never executed silently.

Operating modes

Phase 1 modes (always available)

Mode	When to use
`smoke`	Fast check on main route or recently modified feature
`critical-path`	Core business flows that must always work
`e2e-regression`	Full regression suite
`ajax-heavy`	Pages with async rendering, infinite scroll, filters, hydration
`auth-protected`	Login-dependent flows
`legacy-mpa`	Server-rendered MPA (Blade/ERB/Twig/etc.) with form submit + redirect
`visual-regression`	Screenshot baseline comparison (opt-in)
`perf-budget`	Core Web Vitals budgets (opt-in)
`release-gate`	CI pre-release, rich diagnostics, matrix runs

Phase 2 modes (opt-in via config)

Mode	When to use	Config toggle
`dynamic-scraping`	Data extraction from JS-rendered pages	`scraping.enabled`
`a11y-scan`	WCAG 2.1 AA accessibility check	`axeA11y.enabled`
`mobile-perf`	Mobile-emulated perf budgets	`mobileDesktopMatrix.enabled`
`cross-browser`	Multi-browser execution	`crossBrowser.enabled`
`popup-checkout`	Cross-tab/popup/iframe gateway flows	`crossTabPopup.enabled`

Examples

/playwright-tester mode=smoke
/playwright-tester mode=critical-path files=tests/e2e/checkout.spec.ts
/playwright-tester mode=ajax-heavy folders=tests/e2e/catalog/
/playwright-tester mode=visual-regression tags=@critical
/playwright-tester mode=perf-budget files=tests/e2e/home.spec.ts
/playwright-tester dry-run=true mode=release-gate
/playwright-tester mode=visual-regression update-snapshots=true
/playwright-tester mode=a11y-scan
/playwright-tester mode=cross-browser tags=@smoke

Workflow

After parsing arguments, the command loads the playwright-enterprise-tester skill and delegates to the agent, which follows the standard workflow:

discover → configure → author → execute → diagnose → fix → report → chain

Full specification: skill SKILL.md.

Guardrails

These conditions trigger automatic STOP with user confirmation:

scope=all in local runner mode (dangerous)
update-snapshots=true in ci runner mode (always forbidden)
fix-app-code=true without governance.allowAppCodeChangesWhenExplicitlyEnabled=true
mode=dynamic-scraping on domains not in the scraping.allowedDomains list

Chained skills

Users can configure follow-up skills in .playwright-tester.json:

"skillInvocation": {
  "chainedSkills": {
    "onFrontendChange": "your-perf-review-skill",
    "onPerfBudgetViolation": "your-perf-review-skill"
  }
}

When configured, the agent suggests (or triggers) the follow-up skill after relevant runs. No specific follow-up is hardcoded — the plugin is stack-agnostic.

Configuration

All options are controlled by .playwright-tester.json (repo policy) + env overrides (PWTEST_*). See:

references/runner-modes-local-vs-ci.md
references/targeted-execution-scope.md
docs/CONFIGURATION.md (full parameter reference)

Playwright Tester