26 checks · 9 categories · AI sentiment

What does AI
think
of your site?

Drop your URL. We run 26 checks across performance, SEO, accessibility, security, AI agent readiness, privacy, and more — then ask Claude what it would tell a user who asked about you. Free to start. No login.

https://
No credit card No email gate AI sentiment included
https://acme.com
Audit complete · 27 findings
0
HIGH
TLS 1.0 still accepted
Server completes a handshake at TLS 1.0 — RFC 8996 deprecated this in 2021. Disable on nginx with `ssl_protocols TLSv1.2 TLSv1.3;`.
MED
Hero copy is unclear
Claude couldn't tell what your company does in 10 seconds (confidence 38/100). Headline is vague; CTA reads 'Learn more'.
LOW
No llms.txt found
AI assistants and agents look for /llms.txt as a curated reading list. Without it they crawl HTML inefficiently.
HIGH
Missing HSTS header on HTTPS
Without Strict-Transport-Security, the first request to your domain can be downgraded to HTTP.
24+ more
AI Says · Claude · Citation: 64/100
Built for the AI-first web · audits informed by what ChatGPT, Claude, and Perplexity actually look at
Claude
ChatGPT
Perplexity
Lighthouse
WCAG
Schema.org
GDPR
How it works

Three steps.
No nonsense.

Most audit tools give you a 200-page PDF that nobody reads. We give you a verdict, an AI sentiment read, and a fix list with code snippets. That's it.

— STEP 01

Drop your URL.

No login. No card. We don't want your email — we want your site.

POST /audit
url: https://yoursite.com
→ queued in 0.2s
— STEP 02

We run 26 checks across 9 categories.

Performance, SEO, accessibility, security, email/domain, social, conversion, AI agent readiness, and privacy. Plus a multi-LLM sentiment pass.

✓ TLS protocol + cert + ciphers
✓ Lighthouse (mobile + desktop)
✓ SPF / DMARC / DKIM / CAA
✓ AI agent readiness signals
✓ Claude semantic + perception pass
— STEP 03

Read the verdict.

Overall grade up top. Quick wins, biggest impact, fast fixes. Then every finding with severity, effort, and a copy-pasteable fix. Free unlocks the top 4. Full report unlocks all 27+.

Score: 64/100  ·  C−
27 findings · 9 categories
Quick wins → Biggest impact → Fast fixes
Top-5 prioritized fix list
The methodology

We're
not guessing.

Every audit runs through 25 deterministic checks (performance, security, schema, DNS, TLS, headers, accessibility) plus 2 LLM-powered passes — one for hero-copy clarity and one for the cross-provider sentiment read. Severity, effort, and impact are all on the finding.

↳ Lighthouse 12.x (self-hosted Chromium)
↳ axe-core, openssl, dns_get_record, JSON-LD parser
↳ Claude Sonnet for semantic + perception passes
↳ Crawl up to 50 internal links per audit

PERF Lighthouse + Cache + Compression headers Core Web Vitals
SEO Schema, sitemap, robots, meta tags Schema.org spec
A11Y Headings, alt text, mixed content, hreflang WCAG 2.1 AA
SEC TLS protocol + cert + cipher, SRI, headers Mozilla TLS guide
EMAIL SPF / DMARC / DKIM / CAA RFC 7208 / 7489
SOCIAL Open Graph + Twitter Card + link preview OG protocol
AI llms.txt, Content-Signal, answer-layer Cloudflare playbook
PRIVACY GDPR signals + third-party tracker inventory gdpr.eu checklist
AI·LLM Claude semantic + perception (paid only) Tools API
What you get

The free report, previewed.

Nine categories. Real scores. Concrete remediation guidance with the exact header or code snippet to add. The free version unlocks the top 4 findings; the {{ audit_price_label }} report unlocks all 27+.

Pricing

Free, then
fairly priced.

Most teams run the free audit, fix the obvious things, then upgrade to the full report when they're ready to ship the long tail. Agencies skip both and go straight to white-label.

Free audit
$0

Drop a URL, get a verdict. Best for finding out if you have a problem.

  • Top 4 findings unlocked

  • Scores across all 9 categories

  • Severity + effort estimate per finding

  • Shareable link, expires in 90 days

Run free audit
Agency white-label
$299/mo

Sell audits as your own. 50/mo included. Most agencies bill clients $300–$800 per report.

  • Your logo, colors, and domain

  • 50 audits/mo, $4 each thereafter

  • Company branded PDF

  • Priority queue + API access

Agency details
For agencies

Sell our reports
under your name.

Agencies charge clients $300–$800 per audit. Our cost: $10. The math is obvious. White-label Audora.One for $499/mo and run as many as you need.

Northstar Digital
AUDIT #2104
https://clientsite.com
Quarterly site audit
0
PERFORMANCE
78
SEO
83
Prepared for client by Northstar Digital
Things people ask

FAQ.

How is this different from Lighthouse?

Lighthouse measures performance, accessibility, SEO, and best-practices on one page at a time. Audora runs Lighthouse as one of 26 checks, then layers in things Lighthouse doesn't touch — DNS / TLS / SPF / DMARC / DKIM / CAA, structured data coverage, third-party tracker inventory, GDPR signals, AI agent readiness (llms.txt, Content-Signal, markdown negotiation), and an LLM perception pass that asks Claude what it would tell a user who asked about your site. The free tier surfaces the top 4 findings; the full report surfaces all 27+.

What does the AI actually do?

Two things, both on paid audits. (1) **Hero clarity** — extracts your H1, tagline, and CTA labels and asks Claude to infer in one sentence what the company does, with a 0-100 confidence score. Low confidence = your hero isn't doing its job. (2) **AI perception** — gives Claude (or ChatGPT, Perplexity if configured) the page signals and asks for a structured analysis: identity, content summary, technical signals, trust + privacy, sentiment + credibility + red/green flags, and a TLDR / best-for / watch-out-for verdict. The result renders side-by-side at the top of the report — one card per provider so you see how each LLM sees you.

What's "AI agent readiness"?

It's a category of signals that AI assistants (ChatGPT, Claude, Perplexity, Google AI Overviews) look for when deciding whether to read and cite your site. We check for /llms.txt (the LLM-friendly reading list spec), the Content-Signal directive in robots.txt (~4% of sites have it; explicit ai-train / ai-input / search permissions), markdown content negotiation (server returns text/markdown when the client asks), an answer-layer summary at the top of the page (TL;DR, lede, FAQ pattern), and several emerging well-known endpoints (MCP server card, agent skills index, API catalog, OAuth discovery). Most sites score badly here today, which is the opportunity.

Why does the free report cap at 4 findings?

Because we'd rather give you 4 things you'll actually fix than 40 things you'll ignore. The full {{ audit_price_label }} report is for teams who want the long tail — and yes, it's a real {{ audit_price_label }}, not {{ audit_price_label }}/mo or {{ audit_price_label }} per audit per seat.

Is the {{ audit_price_label }} a one-time charge or recurring?

One-time, per audit. Run it again in 3 months when you've shipped fixes — pay {{ audit_price_label }} again. Most teams audit quarterly. We don't do subscriptions on the report tier; that's the agency plan.

What does white-label include?

Your logo, your colors, your domain, your sender address on the PDF. Reports are generated under audora.one infrastructure but appear to come from your agency. Includes 50 audits/month, additional at $4 each. Most agencies use 30-40.

How long does an audit take?

Free audits typically settle in 30-60 seconds. Paid audits take 2-4 minutes — most of that is Lighthouse running up to 6 passes (mobile + desktop on the homepage plus 4 deeper crawled pages) and the LLM perception calls. The running screen shows real-time per-check progress so you can see what's happening.

Do you store our site data?

We keep the audit results so you can revisit them. We don't keep page content, screenshots, or rendered HTML beyond the audit run itself — only the structured findings, the per-check status, and the LLM perception payloads.

One more push

Find out what AI
thinks of you.

Free audit, no login. 26 checks across 9 categories, plus a multi-LLM perception pass. The first 4 findings are free, and they're usually the ones that matter.

https://