store.sgit.ai / offers / t3
Tier 3 — Can you correct it for my situation?
The vault, with the mandate corrected against your situation rather than against a starting assumption, and a written note of what changed.
| Offer | Tier 3 — t3 |
|---|---|
| Price | £500 |
| When it arrives | 1 to 3 days, from your reply, not from your payment. The clock starts when you send the industry, the use case and the details, because until then there is nothing to correct against. It is said this way round because it is the way round that is true. |
| How it is paid | Payment link. A link, and a printed code beside it. An online payment, so no cross-border rule applies to it. |
| What is true of it today | delivered by a person Done by a named security professional, and done many times: six vaults of this work are published and open to read, including an eleven-step risk-acceptance walk over one evolving register. What you buy at this level is that person's time turned on your situation rather than a template's. |
Who does it
Founder of the sgit.ai network, RiskMandate.ai, The Cyber Boardroom, MyFeeds.ai and VoiceDebrief.ai; former OWASP Board member and organiser of the OWASP Summits; creator of the O2 Platform.
Runs the correction at £500 and the two sessions at £1,500.
One name today, and the list is built for more — other security professionals are being approached to take some of this work. Everybody who runs a review here.
What arrives
- Everything in the vault level, with the mandate corrected against what you told us.
- The delta recomputed against your corrected mandate, with the barrier on every row.
- A written note of what was changed and why, so the correction is checkable rather than trusted.
What this is not, and will not become
- Not a call. Nobody interviews you at this level: you run the prompt where the agent runs and send back what it writes, and the work is done from that.
- Not signed by a named professional. That is the level above.
- Not a compliance assessment and not a risk score. An Agent Behaviour Policy describes and does not judge.
What is taken, and when
| On the order | £100 — a fifth |
|---|---|
| On delivery | £400, invoiced when the work is in your hands |
This level is a person’s time rather than a file, and a deposit is how that is sold honestly from a card: neither side carries the whole amount before anybody has done anything. The split belongs to the offer and not to the rail — a card tapped at the stand takes the same deposit as a link on a phone.
What you do
Nobody needs access to your environment and you send us no credentials. You already have the prompt — MAP-A-GRANT.md ships in every template zip and in every vault, so it is in your hands before you pay as well as after. You run it where the agent runs and send back what it writes. The mandate is corrected against that.
| 1 | You buy this level. Nothing is scheduled and nobody waits. |
|---|---|
| 2 | Within 24 hours a person follows up, and confirms where to send the files. |
| 3 | You paste MAP-A-GRANT.md to the agent, in the environment it actually runs in. It reads its own configuration and prints a table. |
| 4 | You send back the two files it writes — grant.json and mandate.json — and the session record, with no secret in them and your order reference on the message. No credentials, no logs, no access. |
| 5 | The corrected vault comes back, with a written note of what changed and why. |
Step 5 is estimated at one to three days from your reply — from your reply rather than from your payment, because until the details arrive there is nothing to correct against. It is an estimate rather than a measurement, and it is labelled as one; the constraint behind it is one person’s calendar. The 24 hours is riskmandate.ai’s commitment to follow up and is kept on their pages.
The result is computed on your own machine and sent back in bands rather than as raw counts, because a connector list on its own is identifying. The prompt reads and prints; it does not act, and its last line says so. The whole of it is on the page that explains buying, and again on your own page after you pay.
How to buy this
The payment link has not been issued yet
The code on the card redirects here, and this page says what arrives before anything is paid. One price means one standing payment link, printable on a card. The two rails, and why they never meet.
Everything on this page is model generated unless it says otherwise, and it is marked as such where it is delivered. Nothing here is a compliance assessment and nothing here is a mark of conformity — neither word applies and neither appears on the deliverable. Where a finding reaches you, it has been reproduced first: recall-optimised agents run at 0.388 precision, so what is sold is triage and never raw findings. measured 10 Sep 2026
Which agent do you run? → How buying works → Every claim, with its state →