Skip to content
Page audits by Kandr

Every problem flagged.
Every fix already written.

Beacon audits your pages for accessibility, content, and discoverability against a versioned policy library — then shows the specific fix for each finding: the copy to paste, or the markup to change.

WCAG via axe-core SEO + Open Graph Consent & trackers Your brand voice

A Chrome side panel for the page you're on, and a portal at beacon.kandr.io/app for the whole site.

beacon · side panel
northwind.example/pricing audited 4s ago
Accessibility71
Content92
Marketing66
Privacy48
UX88
Findings7
Tracker fires before consent critical privacy · code fix
Image button has no accessible name critical a11y · wcag 4.1.2 · code fix
Meta description is 212 characters warning marketing · copy change
Copyright year reads 2023 warning content · copy change
6
policy areas
0–100
score, banded per page
2
fix kinds: copy & code
1 click
finding to ticket
Features

An audit that ends in a change, not a report

The tools Beacon replaces stop at the flag: “image missing alt text,” “title too long.” Beacon keeps going — to the sentence you should write and the attribute you should add.

Audit the page you're on

Open the Chrome side panel on any page — production, staging, or behind a login — and Beacon reads it as it actually renders, then scores it in seconds.

  • Runs on the rendered DOM
  • Works on pages behind sign-in
  • Point at an element to audit it

Crawl the whole site

Add a site and Beacon discovers its pages from the sitemap, renders each one in a real browser, and re-crawls only what changed.

  • Sitemap discovery
  • Headless browser rendering
  • Incremental re-crawl

The fix, not the flag

Every finding comes with its rendition: a copy change you can paste into the CMS, or a code change against the markup Beacon actually saw.

  • Before-and-after diffs
  • The evidence stored with it
  • Grounded in your repo when connected

Scored, banded, comparable

Each page gets a 0–100 score with a green, amber, or red band, per category and overall — so a product owner knows where to look first.

  • Green from 85, amber from 60
  • Thresholds you can tune
  • Site rollups and coverage

Checks in your own voice

Write your content guide once. Its criteria become judged checks, and your term list becomes exact ones — so “sign in, never log in” is enforced on every page.

  • Voice and tone criteria
  • Preferred and banned terms
  • Per-workspace severities

Tickets with the context attached

Send a finding to Linear or Jira with the evidence, the fix, and the page already in the brief. Beacon checks for an open ticket first so nothing files twice.

  • Linear and Jira
  • Duplicate detection
  • The exact payload recorded

Look inside

What a finding looks like when it arrives

Three findings from one audit, as the side panel shows them. Each names the problem, shows the evidence, and hands over the change.

<button class="icon-btn">
<button class="icon-btn" aria-label="Open cart">
  <img src="/cart.svg" alt="">
</button>

A screen reader announces this as “button” and nothing else.

Code fix · accessibility The attribute, on the element it belongs to Mapped to the WCAG criterion it fails, with the selector so a developer can find it.
Northwind offers a comprehensive suite of solutions designed to help teams of every size streamline workflows, unlock insights, and drive…
Plans for teams of 5 to 5,000. Start free, upgrade when you need SSO and audit logs.

212 → 86 characters. Search results cut off past ~155.

Copy change · marketing The sentence to paste, not a character count Written from what's on the page, in your content guide's voice when you have one.
FindingTracker before consent
DestinationLinear · WEB
Open duplicatesnone
Briefevidence + fix + URL
Handoff · privacy One click to the backlog, with the brief written The ticket carries everything the engineer needs, and Beacon keeps the payload it sent.

How it works

From an open tab to a merged fix in four steps

The finding you see in the side panel is the same record the portal scores, the crawl re-checks, and the ticket links back to.

Open the side panel

On the page you're reviewing, click Beacon. Grant access to that site once and it stays granted.

Run the audit

Beacon reads the rendered page — text, markup, contrast, layout shift, consent behavior — and checks it against the policy library.

Take the fix

Copy the replacement text, or hand the code change to whoever owns the template. Dismiss what doesn't apply, with a reason.

Watch the site

Add the site in the portal and the crawler keeps every page scored, so a regression shows up before a customer finds it.


Policy library

Six areas, one versioned standard

Every check has an id, a severity, and a version, so a score from last month means the same thing as a score from today. Turn checks off, tune severities, or add your own through the content guide.

Accessibility wcag

axe-core rules mapped to WCAG success criteria, plus contrast, alt text, and screen-reader tools.

  • button-name
  • image-alt
  • color-contrast
  • link_new_window

Content copy

The things that slip through review: placeholder copy, template tokens, thin pages, stale years.

  • lorem_ipsum
  • template_token
  • thin_content
  • stale_copyright_year

Marketing seo

Titles, descriptions, canonicals, Open Graph, and structured data — how the page looks everywhere but the page.

  • title_too_long
  • canonical_missing
  • og_image_missing
  • noindex_present

Privacy consent

Measured at runtime: what sets cookies and loads trackers, and whether it waited for consent.

  • cookie_before_consent
  • tracker_without_banner
  • no_policy_link

UX runtime

Layout shift, legible type sizes, and line height, measured as the page actually runs.

  • layout_shift_poor
  • font_below_minimum
  • line_height_too_tight

Brand yours

Your content guide's criteria and term list, composed into checks that run alongside everything else.

  • voice criteria
  • preferred terms
  • banned terms

Built for real sites

It reads the page you ask about. Nothing else.

The full detail is in the privacy policy. The short version:

Only pages you audit

Beacon never records the address of a page you didn't audit and never reads a page's content unless you run an audit on it. No browsing history.

Workspaces stay separate

Sites, findings, guides, and dismissals belong to a workspace, and permissions are enforced on the server. Tracker and repo tokens are stored encrypted.

Not for sale, not for training

Page content is used to produce your findings and nothing more. It isn't sold, isn't shared with ad networks, and isn't used to train models.


Questions

What teams ask before the first audit

Still unsure? Ask us directly — Kandr audits its own sites with Beacon, so the answers come from using it.

How is this different from Lighthouse or an accessibility checker?

Those tools tell you what failed. Beacon covers the same ground — and content, privacy, and brand checks they don't — and then writes the change: replacement copy, or the markup edit on the element it found.

Does it work on staging or pages behind a login?

Yes. The side panel audits whatever is open in your browser, including pages you're signed in to. The crawler covers public pages it can reach from a sitemap.

Do I have to fix everything it finds?

No. Dismiss a finding with a reason and it stays dismissed for that page. Workspace admins can switch checks off or change their severity for every site at once.

Can it check against our own style guide?

Yes. Add your content guide's criteria and a term list in the portal. Criteria become judged checks, terms become exact ones, and both score alongside the built-in library.

What does the extension need permission to do?

It asks for access to a site the first time you audit a page there, and only reads a page when you run an audit. It doesn't run in the background on pages you browse.

Ready when you are

Audit one page. Ship one fix today.

Sign in to the portal with your work account. If your team isn't set up yet, request access and Kandr will create your workspace.