Full Page Test

Use this skill when the user asks to fully test a page, test all navigation paths, check for broken links, verify no JS errors site-wide, run a comprehensive page audit, or crawl and test all routes.

What it does

•Crawls all internal links starting from a base URL (homepage by default)
•Visits each unique page with Playwright (real browser)
•Collects JS errors on every page (SyntaxError, TypeError, ReferenceError, uncaught exceptions, console errors)
•Checks HTTP status of every navigated page (detects 404s, 500s, blocked resources)
•Verifies key elements render on each page (no blank pages)
•Reports a pass/fail summary with details per URL

Quick start

Option A: Playwright spec (recommended)

bash

npx playwright test tests/full_page_test.spec.ts --project="Desktop Chrome"

Or test against the remote site:

bash

VERIFY_REMOTE=1 npx playwright test tests/full_page_test.spec.ts --project="Desktop Chrome"

Option B: Standalone Node.js runner (no Playwright config needed)

bash

node .cursor/skills/full-page-test/scripts/run-full-page-test.js

With options:

bash

node .cursor/skills/full-page-test/scripts/run-full-page-test.js --url https://findtorontoevents.ca --depth 2 --timeout 30000

Configuration

Env / Flag	Default	Description
`FULL_TEST_URL`	`http://localhost:5173`	Base URL to crawl
`FULL_TEST_DEPTH`	`2`	Max link-follow depth (0 = homepage only)
`FULL_TEST_TIMEOUT`	`15000`	Per-page navigation timeout (ms)
`FULL_TEST_SAME_ORIGIN`	`true`	Only follow same-origin links
`FULL_TEST_IGNORE`	(empty)	Comma-separated URL substrings to skip

How it works

Phase 1: Discover links

•Load the base URL, extract all <a href="..."> links
•Filter to same-origin internal links (unless FULL_TEST_SAME_ORIGIN=false)
•Normalize and deduplicate URLs
•Skip anchors (#-only), mailto:, tel:, javascript: links

Phase 2: Visit each page

For each discovered URL (breadth-first up to FULL_TEST_DEPTH):

•Navigate with Playwright (waitUntil: 'networkidle' with timeout fallback to 'domcontentloaded')
•Collect all pageerror events and critical console.error messages
•Check HTTP response status
•Verify document.body has content (not blank)
•Discover new links on this page (if depth allows)

Phase 3: Report

•PASS: All pages returned 2xx, zero JS errors, all pages rendered content
•FAIL: Lists each failing URL with: HTTP status, JS errors, blank page flag

JS error patterns detected

•SyntaxError / Unexpected token
•ReferenceError / TypeError
•ChunkLoadError / Loading chunk
•denied by modsecurity
•Any uncaught exception via pageerror event

Ignored (not failures)

•React hydration mismatch (#418)
•Third-party script errors (Google Analytics, ad scripts)
•favicon.ico 404

Integration with existing tests

This skill complements the existing no_js_errors.spec.ts (single page) and verify_remote_site.spec.ts (remote homepage). Use full-page-test when you need to verify all navigation paths, not just the homepage.

When a test fails

Failure	Likely cause	Fix
JS error on homepage	Chunk not serving JS	Use fix-toronto-events skill
404 on a nav link	Broken link in index.html or Quick Nav	Use fix-nav-menu skill; verify link href
Blank page on sub-route	Missing HTML or server misconfiguration	Check that the path exists on server / in local files
JS error on sub-page only	Page-specific script issue	Inspect the error; check the sub-page's own JS
Timeout on a page	Slow server or infinite redirect	Increase `FULL_TEST_TIMEOUT`; check for redirect loops

Files

•Playwright spec: tests/full_page_test.spec.ts
•Standalone runner: .cursor/skills/full-page-test/scripts/run-full-page-test.js
•Reference docs: reference.md