WebauditOS

The regression-only alerts plus scan comparison are the part I'd actually keep open week to week. Most audit tools dump a fresh 100-item report every run, so the useful signal (what broke since the last scan) gets buried. Pairing a confidence level with raw evidence on each finding also makes it easier to defend a flag when a client asks why a security header or WCAG item matters. One thing I'm not sure about from the feature list: can the REST API trigger a scan on deploy and return just that delta, or is comparison only in the UI?

Showing the evidence and confidence level behind each finding is a smart call, since a bare score is hard to defend to a client. I'd be curious how you handle pages behind a login, as those are often where accessibility and security issues hide. Would a simple authenticated-scan option fit your roadmap?

A confidence level next to each finding is a good idea, since audit tools often flag things that turn out to be false positives. For the GDPR pre-consent check, does it scan from an EU location, and does it also click through the cookie banner to compare what loads after consent?

Great to see tools like WebAuditos making website auditing more accessible. Identifying technical issues, improving performance, and finding SEO opportunities can make a real difference for website owners and marketers. Looking forward to seeing how this product evolves! 🚀

Lab Core Web Vitals can swing a few hundred milliseconds between two runs of the same page. For the regression-only alerts, is there a threshold or a repeat-scan rule so one noisy LCP reading doesn't fire an alert, or does any drop count as a regression?

Showing the raw evidence and a confidence level for every finding is exactly what most audit tools miss — a bare score doesn't tell you what to fix first. The regression-only alerts also seem genuinely useful for monitoring: nobody wants a ping for every tiny fluctuation. I'm curious how the 0–100 score weights the 9 areas — is a security failure weighted more heavily than, say, a UX nitpick?

Showing the raw evidence for each finding is the right call, a bare score is hard to act on. How do you weight the 9 areas in the 0 to 100 score? For a small marketing site, GDPR and accessibility matter more to me than browser compatibility, so being able to see or adjust the weighting would help.

Evidence over a score is the right call. A score makes people chase a number. Do you rank the findings by likely impact, so a small team knows the one thing to fix first this week?

The per-finding evidence would help a small team turn an audit into a concrete fix list. How do you handle multilingual sites: can we compare the same page across locales and catch missing hreflang links or accessibility regressions that only affect one translation?