> For the complete documentation index, see [llms.txt](https://docs.sectoral.xyz/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.sectoral.xyz/ai-agents/agent-wallets.md).

# Wallets for Agents

On Sectoral, software holds an account in its own right instead of borrowing someone else's. We don't treat an agent as a person's account with an API key and a callback URL bolted on. It owns a wallet, carries an identity, and builds a history, all within rules set by its owner. This is how a financial account looks when it is designed with machines in mind from day one.

***

## What each agent receives

Behind every agent account is real infrastructure:

* An ERC-4337 smart account on Robinhood Chain, signed with a key distinct from the parent's
* An Agent ID that lives in the parent's namespace (`@yourname/agent-name`)
* A dedicated encrypted balance inside Sectoral's confidential token contracts
* A payment record that flows into the parent's feed in real time

The agent can hold USDG, receive funding from its parent, and decide on its own when to pay, provided it stays within the limits it was handed. It has genuine economic agency, with a fence around it.

***

## The path from task to settlement

```
Owner sets up the agent account
        ↓
Owner defines a spend policy (limits, approved payees, hours)
        ↓
Owner tops up the agent's vault
        ↓
Agent concludes that a task calls for a payment
        ↓
Policy is checked before any signature happens
        ↓
Within limits: agent signs and submits on its own
        ↓
Over the approval threshold: owner gets a notification and approves or declines
        ↓
Settlement on Robinhood Chain → webhook is sent → parent feed refreshes
```

The step that matters most is in the middle. Policy is enforced by the smart account's validation contract, and it runs before the chain accepts the transaction rather than as an audit after the fact. So an agent isn't just told not to exceed its limits. It simply cannot get a valid transaction onto the chain that would do so. The mechanics are covered in [Spend Policy Engine](/ai-agents/policy-engine.md).

***

## Common uses

* Coding agents that pay for model calls while they work instead of spending down prepaid credit
* Trading systems that operate within a defined risk budget
* Agentic platforms that bill each sub-task almost as it completes
* Developer tools that charge usage directly to the agent generating it, without a billing middleman

***

## Why encryption matters for agent payments

Agents get the same default confidentiality as humans, and the case for it is stronger than it looks at first. Anyone watching a public spend pattern can read a company's usage volume, cost structure and competitive pace as it unfolds. Sectoral encrypts those amounts, while the agent's address and transaction history stay fully visible on-chain like those of any other account. The activity is provable; the numbers stay private.

***

## Read next

* [Spend Policy Engine](/ai-agents/policy-engine.md)
* [Paying with x402](/ai-agents/x402.md)
* [Real-Time Event Stream](/ai-agents/event-stream.md)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.sectoral.xyz/ai-agents/agent-wallets.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.