How the price is calculated, and how to get each discount.
You run a read-only meter on your own system. It counts the code that is yours β discounting third-party libraries, automatically generated files and duplicate copies, item by item.
It writes the result into a file called manifesto.json. That file is what you send us to request the quote β and it is the only thing that leaves your machine.
It is also your proof. Inside it goes a kind of fingerprint of every measured file β a number that changes if a single character of the file changes. So what was quoted and what gets audited are demonstrably the same thing, and you can check it later.
The measurement is free and does not go through us: it runs on your machine, and you send us only the number. Infrastructure products have no measurement β Share is charged per domain, and one application on a server has a single price.
The five products form a ladder, and it climbs on two things at once: how much is examined, and how much access that requires from you.
Share → Code on your machine → Code where it runs → One application on the server → The whole server
The Share is the cheap end of the ladder because the target is one domain on your hosting, not the whole machine: you paste our public key into your panel and the audit runs on our server. It is the cheapest way in β and it already goes inside the server.
In ONE, one agent searches and judges each finding against the code itself. In Deep, several agents search in parallel and different agents judge β without having taken part in the discovery.
Deep costs more and does not always deliver more. It pays off where there are boundaries: separate subsystems, client and server, a binary talking to the host, code talking to the database. In a flat system the gain is small.
We measured both: in a system with many boundaries Deep found 4.5Γ more; in a flat folder, 1.23Γ. So if yours is the second case, we say so. Selling Deep where it does not deliver is the fastest way to lose a client who would have come back.
An audit is not a snapshot, it is a cycle: audit → fix → verify. The verification is what proves the fix worked β and it is usually the part that is missing.
That is why the package is for the same target, and the price is not linear:
| 1 audit | full price |
|---|---|
| 2 audits | 1.7Γ the price of one |
| 3 audits | 2Γ the price of one |
Notice what that means: the third audit costs 30% of the price of one. That is deliberate β we want verifying the fix to be cheap, because the verification is what gives the rest its value.
If you have more than one system, each following audit costs less β as long as not too much time passes between them.
| 1st audit | full price |
|---|---|
| 2nd, within 90 days | 15% off |
| 3rd onwards, always within 90 days of the previous | 25% off |
The 25% is the floor: it does not grow further, and it keeps applying as long as you keep the interval. There is no plan, no monthly fee and no signed commitment β only the 90-day window.
The discount applies up to the value of your largest audit in the last 90 days. If the next one is larger, the discount covers the part up to that value and the difference is charged at full price.
An example with round numbers, just to make the arithmetic clear. Your largest audit in the last 90 days cost $3,000, and now you want one of $4,500. With a discount of 25%:
And from then on your base becomes $4,500. On the next audit up to that amount the discount covers all of it β another $4,500 one would cost $3,375.
The practical effect is that your ceiling rises along with you: each audit larger than the previous ones raises the base, and from then on the discount covers that whole level.
The two discounts stack: the recurrence one applies to the table price, and the package one on top. Three systems followed with a package of 3 come to roughly half the unit price.
It is a real audit, running on a project of yours β not a demo on sample code. The report is yours, and every finding comes with file and line, the trail from entry point to effect, and what to fix.
There is no size limit β pick the project you want to know about. What limits it is not its size: the search looks for the most serious critical problems and stops at the first confirmations. It is not the whole system; it is enough for you to see how we work.
How to do it: give your name and e-mail, confirm through the link you receive, download the kit and run it. No card. The audit runs on your machine β the code never leaves it.
The measurement is not charged either β it is just the count of how many lines your system has, and it runs on your machine.
Write and tell us what you have. If Deep is not worth it for your system, or if the right product is a different one, we will say so β it is cheaper for us than a client who does not come back.