English
Seges Trust
Menu

Example report

What a report actually looks like

Nothing on this page describes a real website

Every finding, number and quoted excerpt below was written by hand for this page. No site was visited, no request was sent, and no third party was checked to produce it. The address it names belongs to a hostname reserved for documentation and test fixtures, so it identifies nobody. Read it as a sample of the format, never as a judgement about any company.

This is the complete report body rendered by the same code as a real report, not a screenshot of selected highlights. Every finding, excerpt, number and measurement below is invented for this specimen. The layout, severity language and limitation states use the same structures as a real report.

Run this check on your own page

How to read it

The evidence
Each finding names what was observed: the file, the header, the rule id, the measured number. You can go and check it on your own page. Where a secret is involved, only a redacted fragment is ever shown, and even that is withheld until the account holds a current DNS ownership record for the host.
Why it matters
Findings are graded, not listed flat. A key shipped in a public bundle and a missing optional header sit at different severities and say so in words, not only in colour, so the top of the report answers the question of what to do first.
The next step
The roll-up turns the findings into one prompt you can paste into whatever coding assistant built the app, with the raw tool output deliberately left out of it. Then you run the same check again on the same page and compare.
What was not assessed
The report ends by saying where it stopped: pages behind a login, business logic, anything the crawl did not reach, and every question that needs a lawyer or an engineer rather than a scanner. A blank section is never presented as a clean one.

The example report

The page in this example

Page checked
https://fixture.example.com/
Report ID
0e0e0e0e0e0e0e0e0e0e0e0e0e0e0e0e
Checked at

A fictional reading-list app created for this specimen. Every finding, excerpt and timing value is synthetic; no site was visited and no request was sent.

Synthetic example. Not a real assessment.

This page raised flags worth fixing

28 checks run · 9 clean · 19 not clean

fixture.example.com

Larkfield Notes: keep every reading list in one place

https://fixture.example.com/

What to fix first

These three numbers count problems found anywhere in this report. The band above counts something else: how many of a fixed list of checks ran, and how many of those came back clean. One check that came back not clean can raise more than one problem here, and several sections below sit outside that fixed list entirely, so the two totals are not meant to match.

  • 6 critical
  • 6 flagged
  • 16 issues
  1. Critical:button-nameThis is a failure reported by axe, an automated accessibility checker that tests the rendered page against the WCAG guidelines. Each one means some people using the page with a screen reader, a keyboard alone, or low vision hit a specific obstacle here, and the impact word is axe's own grading: critical usually means someone cannot complete the task at all, minor means friction. Automated checks catch only part of what is actually wrong with a page for real users, so treat these as the floor rather than the ceiling. They are worth fixing first because the exact failing elements are listed, which makes most of them a targeted edit rather than a redesign.
  2. Critical:labelThis is a failure reported by axe, an automated accessibility checker that tests the rendered page against the WCAG guidelines. Each one means some people using the page with a screen reader, a keyboard alone, or low vision hit a specific obstacle here, and the impact word is axe's own grading: critical usually means someone cannot complete the task at all, minor means friction. Automated checks catch only part of what is actually wrong with a page for real users, so treat these as the floor rather than the ceiling. They are worth fixing first because the exact failing elements are listed, which makes most of them a targeted edit rather than a redesign.
  3. Critical:Content-Security-PolicyA Content Security Policy is a short instruction your server attaches to the page, listing which places the browser may load scripts, styles and embedded frames from. Without it, if anything ever injects a script into your page (through a comment field, a URL parameter, or a third-party widget whose vendor gets compromised), the browser runs it without question. A policy is the difference between one such mistake being contained and it being able to read everything on the page. You can deploy it in report-only mode first, so there is no risk of breaking the site while you learn what it needs.
  4. Critical:Strict-Transport-SecurityStrict-Transport-Security is a one-line instruction telling browsers to only ever reach this site over an encrypted connection, and to remember that for a set period. Without it, the very first request a visitor makes, typing your domain or following an old http link, goes out unencrypted before your redirect can take effect, and that first request is exactly the one someone on the same network can intercept. Redirects alone cannot close that gap, and this header can.
  5. Critical:Stripe publishable keyA string shaped like a credential, an API key, an access token or a password, was found inside a JavaScript file this page serves. Every visitor downloads that file, so if it really is a live key then it is not secret from anyone who opens developer tools, whatever else is guarding it. Some matches are false alarms (publishable keys, test fixtures, values that are meant to be visible), which is why the redacted excerpt stays inside this report and is never copied out to any other tool. Check whether this one is live, because if it is, the only fix that holds is replacing the key rather than deleting the line.
  6. Critical:Analytics write keyA string shaped like a credential, an API key, an access token or a password, was found inside a JavaScript file this page serves. Every visitor downloads that file, so if it really is a live key then it is not secret from anyone who opens developer tools, whatever else is guarding it. Some matches are false alarms (publishable keys, test fixtures, values that are meant to be visible), which is why the redacted excerpt stays inside this report and is never copied out to any other tool. Check whether this one is live, because if it is, the only fix that holds is replacing the key rather than deleting the line.
  7. Flag:color-contrastThis is a failure reported by axe, an automated accessibility checker that tests the rendered page against the WCAG guidelines. Each one means some people using the page with a screen reader, a keyboard alone, or low vision hit a specific obstacle here, and the impact word is axe's own grading: critical usually means someone cannot complete the task at all, minor means friction. Automated checks catch only part of what is actually wrong with a page for real users, so treat these as the floor rather than the ceiling. They are worth fixing first because the exact failing elements are listed, which makes most of them a targeted edit rather than a redesign.
  8. Flag:Privacy policy linkThe scan read every link on this page looking for one that leads to a privacy policy, the document telling visitors what you collect about them and what you do with it, and found none. Visitors who want to check have no way to, and app stores, ad platforms and payment providers routinely refuse or suspend accounts whose site cannot show one. This check saw only the page it was given, so if the policy already exists elsewhere, the fix is likely one footer link.
  9. Flag:Clickjacking protection (X-Frame-Options or frame-ancestors)Nothing here tells browsers whether other websites are allowed to load your page inside a frame, which is a window embedded in their page. That permits clickjacking, where someone lays your real page invisibly over their own buttons so a visitor believes they are clicking one thing and actually clicks yours. It matters most where there is a logged-in action worth hijacking and much less on a static brochure page, but the fix is a single header either way, so there is little reason to leave it open.
  10. Flag:X-Content-Type-Options: nosniffThis is a one-line instruction telling the browser to trust the file type your server declares rather than guessing from the contents. Without it, a browser can decide that something you serve as plain text or as a user upload is really JavaScript or HTML and run it, which is how a harmless-looking upload becomes executable code. The value is fixed, there is nothing site-specific to decide, and it has essentially no chance of breaking anything, which makes it one of the cheapest items in this whole report.
  11. Flag:Referrer-Policy sends the full URL to every origin (unsafe-url)Referrer-Policy controls how much of your page's address the browser reveals to other sites, and yours is set to a value that hands over the whole thing, path and query string included, to every destination. If any of your URLs carry a password reset token, a search term, a customer id or an order reference, that value is being passed to third parties, including every analytics and advertising script on the page. This is graded above a merely missing policy because the header is present and actively choosing the leaky behaviour, and changing one word fixes it.
  12. Flag:Cookie Secure attributeA cookie is a small piece of data your site asks the browser to keep and send back on each visit, often the token that keeps someone signed in. At least one of yours is set without the Secure attribute, which means the browser is willing to transmit it over an unencrypted connection, where anyone on the same network path can read it. The affected cookie names are listed in the evidence, and this is a single option added wherever each one is set.

16 further finding(s), in the sections below

Hand this to your coding agent

Copies a plain-text prompt describing every failing check on this page, with the exact evidence recorded for each one: the header names and values, the paths that responded, the CSS selectors, the library versions, the cookie names. Paste it into Claude Code, Codex, or whatever you already have open.

Secret-scan findings are never included. Their excerpts are fragments of real credentials, so the prompt states that they were left out instead. Read them here and rotate them yourself.

Messaging & CTA clarity

  • 1 heading(s) found (h1/h2)
  • 1 call-to-action element(s) detected
  • Issue:Flag: no CTA visible in the initial viewport

Accessibility (automated, axe-core)

  • Flag:color-contrast (serious impact) · 2 element(s) · .hero-sub, .footer-note
  • Critical:button-name (critical impact) · 1 element(s) · header > button.menu
  • Critical:label (critical impact) · 1 element(s) · #signup-email

Basic trust signals

  • HTTPS
  • Flag:Privacy policy link
  • Contact path

Presence checks only. This is not a verdict on legitimacy, security, or compliance.

Response security headers

  • Critical:Content-Security-Policy
  • Critical:Strict-Transport-Security
  • Flag:Clickjacking protection (X-Frame-Options or frame-ancestors)
  • Flag:X-Content-Type-Options: nosniff
  • Referrer-Policy
  • Issue:Permissions-Policy
  • Issue:Cross-Origin-Opener-Policy
  • Issue:Cross-Origin-Embedder-Policy
  • Issue:Cross-Origin-Resource-Policy

Header presence on this one response only, not a full security assessment.

Security header values (not just presence)

  • Flag:Referrer-Policy sends the full URL to every origin (unsafe-url) · value applied: unsafe-url

Grades the value each header actually carries on this one response. A header can be present and still opt out of the very protection it names, which a presence-only check cannot see.

Software version disclosure in headers

  • Issue:A response header discloses a specific version
2 identifying header(s) observed on the response
  • Issue:X-Powered-By: Express 4.17.1 · discloses a specific version
  • Server: nginx · names a product but no version, which is not a finding

A precise version number tells an attacker exactly which published CVE list to consult for your stack. A bare product name does not, and is reported here as an observation only.

Cookie security attributes

  • Flag:Cookie Secure attribute
  • Issue:Cookie HttpOnly attribute (or likely intentionally client-readable)
  • Cookie SameSite attribute
  • Issue:lk_session · missing Secure · missing HttpOnly
  • lk_theme (likely intentionally client-readable, not flagged)

Set-Cookie attribute presence on this one response only, not a full session-security review.

CORS configuration

Not checked at this tier: testing CORS means sending an origin the site has no reason to trust, and this report does not do that.

Source-map annotations in scripts

4 same-origin script(s) examined for a source-map annotation.

  • Issue:https://fixture.example.com/assets/app.9f21c4.js · declares a source map at app.9f21c4.js.map

Reports only what each script DECLARES in its own sourceMappingURL comment, read from script bodies already downloaded for the secret scan. No .map file is ever requested, so this is not a claim that your original source is retrievable, only that the bundle points at it.

Common leaked-file exposure

Not checked at this tier: this report does not request paths you have not published, so none of these common file paths were probed. That is not a finding that nothing is exposed.

Email authentication (SPF/DMARC)

  • SPF record
  • Issue:DMARC record

Checks the exact domain you submitted only, not any parent or sibling domain. A missing record means email claiming to be from this domain is easier to spoof for phishing.

TLS / certificate configuration

No issues in this section
  • Modern TLS protocol (TLSv1.3)
  • Certificate is valid and trusted

Checks the negotiated TLS protocol version, cipher, and certificate validity/expiry on the exact hostname you submitted, port 443 only.

SEO fundamentals

  • Title tag (49 chars)
  • Issue:Meta description
  • Issue:Canonical link
  • Issue:Open Graph tags

This is an outside-in read of one page. It cannot tell you whether the page is indexed, where it ranks, or whether any AI assistant cites it, because all three live inside systems only their operators can see. It did not fetch the alternate URLs your hreflang tags name, so reciprocal return links were not confirmed, and it did not call any third-party validation service. Length judgements on the title and description are character-count heuristics: real search results truncate by rendered pixel width, which differs sharply between scripts.

Load timing (single load)

  • Issue:Time to first byte: 1240 ms · Needs improvement
  • Flag:Largest contentful paint: 4380 ms · Poor
  • Issue:Cumulative Layout Shift: 0.19 · Needs improvement
  • DOM content loaded (when the page's HTML finished loading and parsing, before images and scripts finish): 2610 ms

A single-load sample, not a Real User Monitoring (RUM) trend.

Compliance-adjacent link presence

  • Terms of Service link
  • Issue:Privacy policy link · Not linked from this page.
  • Issue:Accessibility statement link · Not linked from this page.
  • Issue:Data subject request / privacy-rights contact · Not linked from this page.
  • Issue:Refund/cancellation policy link (commerce intent detected) · Not linked from this page.

Link/text presence only, not a legal compliance review. See limitations below.

This checks only the single public page you submitted. Policies behind a signup or login wall, or linked only from an in-app footer, are not visible to this check.

Client-side secret scan (same-origin scripts)

4 same-origin script(s) scanned.

  • Critical:Stripe publishable key · /assets/app.9f21c4.js · pk_live_51Nx****************************
  • Critical:Analytics write key · /assets/app.9f21c4.js · wk_****************************

Scans only same-origin scripts already loaded during this page visit. Assets served from a separate CDN host are outside this scope. Pattern-matching only, not a guarantee of no leaked secrets.

Additional pages checked

Each additional page below was checked for accessibility, security headers, and secret exposure only. See Limitations for what this bounded crawl does not cover.

No additional same-origin pages were measured.

Journey reachability

This check also measured same-origin pages linked directly from the submitted page (one level deep).

  • 3 page(s) measured
  • 11 same-origin page(s) found linked from this page

Limitations: Additional pages limited to same-origin links found on the submitted page (one level deep), capped to a small number of pages, Not a security or compliance audit, Presence check, not a legal review

Rendered by the same component as a real report, over invented input. It is not a penetration test, a certification, a legal opinion, or a promise that any product is secure.

Run this check on your own page

See what it costs