GuardScan · PromptGuard · ArkGuard

Find the holes in your website — before someone else does.

GuardScan checks your SSL certificate [your website's ID card] and the 7 security headers [the standard locks every site needs] in 5 seconds — free, no signup, plain-English fixes. Then lock down your AI with PromptGuard and ArkGuard.

Free·No credit card·Report in 5 seconds·1,219 companies scanned
The Ark Warranty: if our tools miss a vulnerability we should have caught, we pay you $500.
Live GuardScan security report for x.ai — 43 out of 100
A real GuardScan report — x.ai, scanned live: 43/100. Four of seven locks missing, with plain-English fixes for each.

Four holes most businesses don't know they have

Security doesn't feel urgent — until one of these becomes your headline.

Your website is missing locks

Missing security headers leave you exposed to data theft, clickjacking [traps hidden on pages you think are safe], and impersonation. We tested 1,219 AI companies — most were missing at least one lock. The fix takes ten minutes, once you know what to look for.

Your AI chatbot gives away secrets

Prompt injection [a cleverly-worded question that tricks your bot] can make it reveal its secret instructions, customer data, or pricing. One question, and your "private" AI becomes an open book.

Your AI agents act without permission

Agents with broad permissions can send sensitive data to the wrong place, delete files, or run commands you never approved — and you only find out after the damage is done.

Your AI-written code ships with blind spots

AI coding tools write code fast — but who checks it? Hardcoded passwords, missing validation, and quietly weakened tests sail straight into production. You find out after the breach, not before.

Four tools that fix it

Each tool closes one of the holes above. Use one, two, three, or all four.

Free

GuardScan

Think of it as a free home-security check for your website. We check the locks most sites forget — your SSL certificate [your website's ID card] and the 7 security headers [the basic locks every site needs] — and tell you, in plain English, exactly what's open and how to close it.

  • Checks all 7 locks every website needs
  • Shows what AI scrapers & bots can read
  • Copy-paste fixes — no jargon
  • Open-source — view on GitHub
Run free scan
$45/mo

PromptGuard

A cleverly-worded question can make your chatbot spill customer data, your secret instructions, or your pricing. PromptGuard attacks your bot the same way a hacker would — 20 different ways — and shows you exactly what it gives away.

  • 20 real attack attempts on your bot
  • See exactly what it reveals
  • Plain-English report + fixes
  • Monthly re-checks as attacks evolve
Choose PromptGuard — $45/mo
$45/mo

ArkGuard

Your AI agent can send emails, open files, and run commands. ArkGuard puts a security guard around it so it can only do what you approve — and pings you the instant it tries anything else.

  • Whitelist what your AI can access
  • Block network, files, or commands
  • Instant alert on any escape attempt
  • Works while you sleep
Choose ArkGuard — $45/mo
$45/mo

ShipGuard

Your AI wrote the code — but nobody checked it. ShipGuard reviews AI-built code with a different AI, the way you'd never let an employee grade their own exam. It hunts hardcoded passwords, missing validation, and softened tests — then hands you a plain-English fix list.

  • Independent cross-model review of every change
  • Catches hardcoded credentials & exposed secrets
  • Flags tests quietly weakened to "pass"
  • $500 warranty on everything we review
Choose ShipGuard — $45/mo

One tool $45 · two tools $63 · all three $81/mosave up to 40%

Testimonials

What people are saying

Trusted by founders and security teams who take website security seriously.

Our first score was 43/100. We fixed six of the seven locks over a weekend — a fraction of the pentest quote we were about to sign.

FounderSaaS company

We fixed our missing headers before a security review with an enterprise buyer. Their first question was: who did this audit?

CTOFintech startup

The report is our onboarding tool now. We send it to prospects before they ask — it shows we take security seriously.

Head of SalesB2B platform

We ran our biggest competitor through it before launch. They’d fail today. That’s a genuine edge.

FounderAI startup
The numbers

Proof, not promises.

1,219
companies scanned
335
passed every check
7
locks checked per site
$500
warranty if we miss something

Only 335 of the 1,219 AI companies we scanned had every lock in place — the sites that take security seriously stand out. See what they lock down →

How it works

Three steps. Five minutes. Zero IT required.

1

Start with a free scan

Enter your name, email, and website. Get your security report instantly — free, with copy-paste fixes.

2

Pick your protection

Choose the tools you need. $45 per tool, or both for $63. One email account. Two-minute setup.

3

Sleep better tonight

Monthly audit reports land in your inbox. Fix guides included. If we miss something, we pay you $500.

See it happen — free

Watch a real site get scanned in front of you, or scan your own. No credit card, no account, no sales call.

Try a live demo

One click. We scan a real, well-known site right now — no email, no typing — and show you exactly what its missing locks leave unprotected.

Scan my website

Your domain, your report. We check your SSL certificate and all 7 security headers — the locks every website should have — and hand you plain-English fixes.

FREE
Scan my website

Run an AI chatbot or AI agents? See how PromptGuard + ArkGuard protect them →

Simple, transparent pricing

One account. Email-based. Choose your tools.

Enter your email, then choose a plan to subscribe.

1 Tool

PromptGuard, ArkGuard, or ShipGuard

$45/mo
  • Any one tool of your choice
  • Email-based account
  • Monthly audit reports
  • Fix guides included
PromptGuard — $45/mo ArkGuard — $45/mo ShipGuard — $45/mo

Both Tools

Save 30% on the pair

$63/mo
  • Any two tools
  • Email-based account
  • Monthly audit reports
  • Fix guides included
Get Both — $63/mo

Free Scan

Comprehensive, no credit card

$0
  • Website security scan
  • SSL + all 7 security headers
  • Full fix report with instructions
  • Repeatable — scan anytime
Start free scan
1 tool = $45  ·  2 tools = $63  ·  3 tools = $81  ·  cancel anytime, no contracts

For companies

One subscription protects your whole company's AI — every chatbot, every agent, every page. PromptGuard + ArkGuard, one admin dashboard, one bill.

Frequently asked questions

What's the difference between the four tools?

GuardScan is the free one — it checks your website's locks (SSL + security headers) and shows what a hacker could see. PromptGuard tests your AI chatbot — it tries to trick it into revealing customer data or secrets, just like a hacker would. ArkGuard watches your AI agent — it makes sure the agent can only do what you've approved. ShipGuard reviews your AI-written code with an independent AI — it catches hardcoded passwords and weakened tests before they ship. Different problems, different tools.

How do I create an account?

Just your email. Sign up, pick your tools, and connect in minutes — GuardScan runs from our website, and PromptGuard + ArkGuard connect to your AI. One email = one account. For companies, your admin adds your people by email.

What's the $500 Ark Warranty?

If our tools miss a vulnerability they should have caught, we pay you $500. No questions asked. Included with the Both Tools bundle and every plan.

Is the free scan really free?

Yes. No credit card. No sales call. You get a full report on your website's security with specific fix instructions — and you can re-scan anytime, as often as you like.

Can I cancel anytime?

Yes. Monthly billing. Cancel anytime from your dashboard. No contracts, no minimum commitment.

What platforms do you support?

GuardScan works with any website. PromptGuard works with any AI bot with an API endpoint (ChatGPT wrappers, Claude apps, custom agents). ArkGuard works with any AI agent running in a container or sandboxed environment [a locked room your AI runs in].

See what a hacker sees.
Before they use it against you.