Site QA console your web team runs before every launch

I want a site QA console my web team opens before we push a redesign or a batch of landing pages live. The point of it is to catch problems ourselves before the client does, so it has to work as a real launch checklist that two people can work through together, not a one shot report that gets read once and forgotten.

The app opens on a sweep setup screen. I enter a client name and a starting URL, then set the scope: keep the crawl to that one domain only, and set a maximum page count so a big site does not run away. On the same screen I set my own thresholds for what counts as a failing Lighthouse score in each category, and the word count below which a page counts as thin. Give these sensible defaults I can change. Starting a sweep calls Browserless Start Crawl with the domain restriction and page cap applied.

While the crawl runs, the app polls Browserless Get Crawl Status and Results and shows live progress: pages discovered, pages completed, and results paging in as they arrive rather than making me wait for the whole thing. The Browserless Crawl API is beta and Cloud plan only, so if a crawl cannot be started on my plan the app must degrade gracefully instead of dead ending. Fall back to Browserless Map Site URLs to discover pages from the sitemap and page links, and also let me paste a list of URLs in by hand. On that fallback path, gather each page's title, meta description and word count with Browserless Scrape Elements. Everything downstream in the app works identically no matter how the page list was built.

The main screen is a table of every discovered page showing its status code, title, meta description and word count. The app flags problems automatically: broken pages, redirect chains, missing titles, titles duplicated across more than one page, and thin content below my word count threshold. I can filter the table down to just the flagged rows, sort by problem type, and search by URL.

Each page can be tagged with a role: homepage, landing page, conversion page, article, or other. Roles are what keep the expensive checks affordable on a large site. Instead of auditing hundreds of pages, I pick one representative page per role and audit that one, and the app should nudge me toward that by showing which roles do not yet have an audited page.

From the table I select those key template pages and run Browserless Run Lighthouse Audit on them, showing performance, accessibility, best practices and SEO scores. Any score below the threshold I set is flagged clearly. Scores are stored against the page so they sit alongside its crawl data in the table rather than living in a separate report.

For the same selected pages the app captures Browserless Take Screenshot at a desktop width and a mobile width, and shows them in a single gallery so I can eyeball every layout in one pass instead of resizing a browser by hand. Clicking any shot opens it full size, and desktop and mobile for a page sit next to each other so a layout that breaks narrow is obvious.

Every page carries a review mark I set myself: Pass, Needs work, or Ignore. Marks persist between sessions and are shared across the team, so two of us can work the same sweep at once and see what the other has already cleared, including who set each mark and when. I can leave a short note on any page explaining what needs fixing. A counter at the top of the screen shows how many pages are still unreviewed, which is what makes this a checklist rather than a dashboard.

An Export button finishes the sweep. It renders the findings with Browserless Generate PDF, covering the flagged pages, the Lighthouse scores against my thresholds, the screenshots, and the review marks with their notes. The app then files that report into a per client folder in Google Drive using Create Folder, uploads it with Upload File (Multipart), produces a shareable link with Create Permission, and posts a short summary plus the link to our team channel with Slack Bot Send a Message. The Slack summary should lead with the numbers that matter: pages crawled, pages broken, pages below threshold, and pages still unreviewed.

Keep past sweeps per client so I can open the previous one before a launch and see what we flagged last time, and so the client folder in Drive builds up a history of reports rather than one file being overwritten each run.