> ## Documentation Index
> Fetch the complete documentation index at: https://docs.walletsuite.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart

> Connect an agent to WalletSuite and run your first read call.

WalletSuite operates the hosted platform — see [Overview](/ai-agents/overview). You connect an agent by authorizing it with its key: nothing to install.

## 1. Authorize your agent

Authorize with your agent's key to get a short-lived token for the connection. The key stays on your server and never reaches the model. See [Connect your agent](/ai-agents/token-exchange).

## 2. Point your MCP client at the hosted endpoint

Attach the short-lived token as an `Authorization: Bearer` header and connect. This works for any framework or custom runtime that speaks MCP over HTTP.

```json theme={null}
{
  "mcpServers": {
    "walletsuite": {
      "url": "https://mcp.walletsuite.io",
      "headers": { "Authorization": "Bearer <token>" }
    }
  }
}
```

## 3. Try it

With the agent connected at the read level, ask:

* `What's the ETH price right now?`
* `Check balances for 0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045`
* `Show me recent transactions for 0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045 on ethereum`
* `Resolve USDC on ethereum so I can prepare a transfer`

## 4. Grant more capability

Move from read to prepare, execute, or governed autonomous by binding a broader tool profile to the agent and requesting the matching scopes at token exchange. See [Execution Levels](/ai-agents/execution-levels).

## Key concepts

<CardGroup cols={3}>
  <Card title="Agents & Authority" icon="user-shield" href="/core-concepts/agents-and-authority">
    The five bound things that make an agent, and why the model never authorizes.
  </Card>

  <Card title="Policy Gates" icon="shield-check" href="/core-concepts/policy-gates">
    Deny-by-default allowlist rules, evaluated before any share is released.
  </Card>

  <Card title="Structured Errors" icon="message-exclamation" href="/core-concepts/structured-errors">
    Category, code, and required action on every error. Agents recover programmatically.
  </Card>
</CardGroup>

## Where to go next

* [Connect your agent](/ai-agents/token-exchange) — the full credential-exchange flow
* [Choose your surface](/ai-agents/choose-your-setup) — admin, embedded, customer-owned, background, or deterministic
* [Execution Levels](/ai-agents/execution-levels) — from read-only to governed execution
* [Tool Reference](/ai-agents/tool-reference) — the complete tool surface