AuditLume library
Turn automated observations into careful, concrete reviews. These guides explain what the scanner sees, what needs human judgement and how to choose a sensible next action.
Understand how titles, identity, contact routes and transparent wording support informed decisions.
Compare the page title and opening promise with what the organisation actually offers. Check that a visitor can identify the publisher, find a suitable contact route and understand the next step without guessing.
Use automated signals as a starting point, then complete the review with keyboard, screen and content checks.
AuditLume can inspect the declared language, alternative text, form labels, viewport, H1 count and a simple heading signal in initial HTML. These checks expose useful omissions, but they cannot establish that a page works for every person or assistive technology.
Separate detectable links and tracker signals from the legal and contextual decisions that require specialist review.
The scanner looks for a visible privacy link and known tracker patterns in the initial page. It can flag a tracker when it finds no common consent signal, but it cannot determine the lawful basis, verify every subsequent request or judge whether policy wording fits a particular organisation.
Interpret HSTS, CSP and protective response headers as observable layers rather than proof that a site is secure.
AuditLume reads headers from the final public response. On HTTPS it checks HSTS and also looks for Content-Security-Policy, X-Content-Type-Options and Referrer-Policy. Their presence can reduce particular browser risks; their absence does not prove exploitation, and their presence does not rule out vulnerabilities.
Review title, description and main heading so people and search systems can understand each page more easily.
The automated review checks whether the title is present and reasonably descriptive, whether a meta description exists and how many H1 elements appear in initial HTML. These are diagnostic signals, not ranking rules. Useful content, crawling and indexing depend on much more than this short list.