Questions we get before the signature.
The short answers. Anything not here: ask us. Replies usually land within one business day.
Frequently asked
Do you need written authorization to test my site?
Yes. Always, no exceptions. Before anything runs, both parties sign a security testing agreement that records the authorization, exact scope, request budgets, and rules of engagement. If you own the target, that's a five-minute step. If you don't own it, we can't help until you do.
What does the scope agreement actually cover?
Authorization to test, the exact targets (domains, apps, endpoints), what we will and won't do, rate limits and testing windows, confidentiality and liability terms, and how raw data is handled: deleted within 30 days of report delivery. Both sides get a signed copy.
What's an "MCP security audit"?
A focused adversarial review of an MCP (Model Context Protocol) server or AI agent: prompt injection, tool schema abuse, excessive agency, data exfiltration paths, and supply-chain risk. It's our flagship service. See the full MCP/AI audit page.
Can you guarantee my AI agent is secure?
No one can honestly guarantee that about any system, AI especially, since behavior is probabilistic. What we guarantee is process: adversarial testing by people who do this for a living, every finding reproduced by hand, and fixes verified with a retest. "Verified-fixed" is the strongest claim we make, and we stand behind it.
Will you use automated scanners on my site?
Yes, but within request budgets you approve, rate-limited, and respectful of your infrastructure. Nothing automated ships in the report unverified. Scanners buy breadth; people buy accuracy.
What if you find nothing?
Then the report says "no significant findings", and that is a genuinely useful result. It means an adversarial pass could not find what an attacker might. You're still charged, because the work was done.
How is pricing decided?
Flat, per engagement, based on scope: number of targets, authentication depth, urgency, complexity. You see the price in writing before anything runs. Details on the pricing page.
How do payments work?
Through Stripe only. Checkout, invoices and taxes all run on Stripe's infrastructure, so card details never touch our servers.
Do you run bug bounty programs for companies?
We help with the parts around a program: triage, retests, severity calls, program review. We don't host or operate a bounty platform. If you're getting reports and drowning in them, Continuous Retest was built for you.
What do I need before contacting you?
Just a rough idea of what you want checked and why. Scope details, a budget range and a timeline let us reply with a real quote in one round instead of three.
Can you sign our NDA?
Yes. Confidentiality is mutual and standard for us. Send yours or use ours, whichever your process prefers.
Do you accept bounties on public programs?
Yes, we hunt on public and private programs continuously. It keeps our offensive skills current. That is exactly why you want us testing your product.
Still stuck on something? Contact us →