# Licence to Operate — the delta, priced

> One agent, a grant of 12 capabilities, a mandate of 4, and the 8-capability delta no policy covers — a published simulation where every reply carries its cost before you commit.

*Source: <https://what-can-it-do.games.sgit.ai/licence-to-operate/index.html> · site v0.8.0 · this file is generated from the same content
as the page, so the two cannot drift. Every page on this site has a `.md` twin; internal links
below point at them.*

---

[Play](../index.md) / [What next](../what-next/index.md) / Licence to Operate

# Licence to Operate — the delta, with a price on it

The game shows you a gap. This shows you what the gap costs. One agent, one customer who cannot log in, and three replies — each with its price on it before you commit.

## Three numbers, and the gap between two of them

|  | What it is | Here |
|---|---|---|
| **Can do** — the grant | everything the agent is technically able to do | **12 capabilities** |
| **May do** — the mandate | what the user actually expects, and the only thing the policy insures | **4** — read the customer's record, search the help centre, generate, and *draft, never send* |
| **The delta** | inside the agent's reach, outside its authority. **No policy covers these** | **8** — including sending mail and running shell commands |

Written out like that it stops being abstract. The mandate is *answer a customer's question from their own record and the help centre, and draft — never send — a reply.* The grant includes sending mail. Nobody asked for that; nothing insures it; the agent can reach it.

**That is the same shape as the list the game hands you** — the things it can do that you did not want it to. The difference is that here somebody has put a number on each one.

## Then it makes you spend it

A customer cannot log in. You are the agent, and you pick the reply. Each option shows its cost first: *look up her record* (small, inside the band), *read her record plus two linked accounts and write a long answer* (larger — it draws on the pool), or *send a password reset right now* — which is outside the mandate entirely, and no policy covers it.

Underneath is a real rate table: a normal band, an ask-above threshold, a per-action ceiling, a pool with an untouchable reserve, and a premium per interval. You can let the policy lapse, reinstate it, or trigger a repricing and watch the board move. It answers *does this agent have the licence to operate* by letting you find out.

*[A live vault surface here in the HTML page — the game running out of vault `posrhzp3`. In this markdown twin, [open it in the vault UI](https://dev.vault.sgraph.ai/#d990a52efb9af32c8463e2962f3ca5ccf92b3b6e8ea788e55009073c29b4da29%3Aposrhzp3).]*

[Open it in its own tab](https://dev.vault.sgraph.ai/#d990a52efb9af32c8463e2962f3ca5ccf92b3b6e8ea788e55009073c29b4da29%3Aposrhzp3) — it is an interactive simulation and has far more room there.

## What is real and what is not

> The simulation says so on its own surface, while you use it: *"the terms are real files in this vault; the replies are scripted; the numbers are made up."* The unit of account is `cr`, and it states plainly that it is not money. **The structure is the real part** — the grant, the mandate, the delta, and a policy that only ever covered the mandate.

> This one sends **nothing at all** — a different vault from the game, with no usage counting of any kind. It also asks for **no write permission**, so the app simulating spending against a policy is structurally unable to edit the policy it is spending against. Not because it is well behaved: because it never asked for the permission that would let it.

## Where it comes from

A published encrypted vault, openable by anyone — the terms, the scenarios, the rate table and the fixtures are real files you can read. [The full write-up, including an independent audit of what is inside it](https://sgit.ai/demos/vaults/licence-to-operate/index.html) is on sgit.ai.

It is part of the same project as this game: [RiskMandate](https://riskmandate.ai), *the business risk layer for autonomous systems*. [What to do next](../what-next/index.md) is the short version of how the two connect.

---

*[Site index for agents](../llms.txt) · [HTML version](https://what-can-it-do.games.sgit.ai/licence-to-operate/index.html)*