Hullwise vs Mozilla Observatory

The MDN HTTP Observatory, formerly Mozilla Observatory, mainly checks HTTP headers and related settings such as cookies, the redirect to HTTPS and SRI, with a score and a grade.

Hullwise requests the page once and reads security headers, certificate, email protection, software versions, the served HTML and the JavaScript files the page includes. Below is where that overlaps with Mozilla Observatory.

https://
01 · Checks

Checks side by side

Where Mozilla Observatory does more, the table says so and highlights it in green. Here that applies to: Security headers.

CheckHullwiseMozilla Observatory
Security headersYes, six headers checked for presence, plus server and technology disclosureYes, the core feature, with points per test. Also checks CORS and Cross-Origin-Resource-Policy
Mozilla Observatory does more here
TLS and certificate in depthBasics: valid, expiry date, TLS 1.0 or 1.1. No review of cipher suitesNo. The TLS and certificate evaluation was removed in 2024. It checks whether the site redirects to HTTPS first
SPF and DMARCYes, via public DNS, plus MTA-STS or DANE and the limit of 10 SPF lookupsNot mentioned
Versions with known vulnerabilitiesYes, detected JavaScript libraries live against OSV.dev, prioritized by CISA KEV (actively exploited) and FIRST EPSS, plus end of support via endoflife.date and insecure WordPress versions via WordPress.orgNo, not part of the test according to MDN
Technology detectionYes, with the rules of webappanalyzer (a fork of the Wappalyzer rules) and Retire.js, plus platforms and hosts such as Lovable, Bolt, v0, Replit, Vercel, Netlify and SupabaseNot mentioned
Secrets in the HTMLYes, the same patterns as in the JavaScript files, also in embedded data such as __NEXT_DATA__, plus Google API keys as a notice. We report the type and where it was found, the value only shortenedNot mentioned
Keys in the JavaScript files of vibe-coded appsYes, 16 types of our own for secret keys and credentials, such as Supabase service_role, OpenAI, Anthropic, Stripe and AWS, plus over 100 providers from the gitleaks rules. We read up to 50 files from the same domain: included files, parts loaded when switching pages, for Next.js the parts of every page, and configuration files such as /config.json that the code fetches on startup. We report the type and where it was found, the value only shortened, for example sk_live_…9f2A, never in fullNot mentioned
Third-party scripts without SRIYesYes
CookiesYes, Secure and HttpOnlyYes, secure cookie settings
Malware and blocklistsNoNot mentioned
Explanation of findingsOne or two sentences per finding in English or German, with a link to the fix guide. Without a paid plan, two findings come in full, all of them with BusinessScore of up to 145 and a grade from A+ to F
APIYes, with a free key 100 checks per hour and a shortened report, full with Business. No REST API without a key, only MCPYes, v2 API. A site can be rescanned at most every 60 seconds
MCP for AI agentsYes, for Claude, ChatGPT, Cursor, Codex and VS Code. Without a key 20 checks per hour and a shortened report, full with BusinessNot mentioned
Weekly monitoringYes, with Business up to 10 domains, with an email as soon as a grade dropsNot mentioned
Manual review (audit)Yes, from €1,500 per application, only after proof of ownershipNot mentioned

As of October 2026. "Not mentioned" means the Mozilla Observatory pages linked below say nothing about it. The feature may still exist there. What Hullwise checks in detail is described under How we check.

02 · The alternative

When Mozilla Observatory is the better fit

  • You only want to check your headers, with points per test. That is what the Observatory is built for.
  • You want CORS and Cross-Origin-Resource-Policy included. We don't check those.
  • You already work with the MDN documentation and want the explanations right there.

You can use both tools side by side. Hullwise gives you an overview across six areas; for a single area, Mozilla Observatory takes a closer look.

03 · AI agents

What your AI agent can do with it

Hullwise has an MCP server. Add it to Claude, ChatGPT, Cursor, Codex or VS Code, and your agent checks sites itself with scan_website and shows what is checked with list_checks. Over MCP, you get 20 checks per hour without a key and 100 with a free key. The REST API always requires a key. Without a paid plan, the API and MCP return two findings in full; with Business, the complete report.

The Observatory has an API that returns grade and score. The MDN pages do not mention an MCP server.

According to MDN, the scan history of every domain is public. Hullwise results are not publicly visible.

Set up MCP

04 · Next steps

What happens after a finding

The Observatory shows score and grade per test. The MDN pages mention no fixing service and no review by people.

At Hullwise, every finding leads to a guide on how to fix it. With Business, we then check up to 10 domains every week and email you as soon as a grade drops (€149 per month). The Security Audit is a manual review that starts only after you prove ownership. It also covers what is not visible from the outside, such as open databases and access permissions. It starts at €1,500 per application.

See pricing

More comparisons

All trademarks belong to their owners and are not affiliated with Hullwise. Information about Mozilla Observatory is based on its own pages, as of October 2026.

Is your app
watertight?

Hullwise finds the leaks before someone else does: exposed keys, database credentials and outdated software. Every finding comes with steps to fix it.

https://

Click a finding to mark it as fixed.

See pricingHow a scan works

Ship fast. Stay hullwise.