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

# Quickstart

Connect your agent and let it buy what it needs, within limits you set. Two routes: no code in your AI app, or via the CLI.

## Route A: In your AI app (no code)

{% hint style="info" %}
Live on staging only, coming soon to production
{% endhint %}

**1. Add the vAPI connector**

In Claude Desktop or another AI app that supports custom connectors, add the hosted vAPI connector. 

```
https://mcp.vapinetwork.ai/mcp
```

Discovery (searching the catalog) is free and needs no account.

**2. Set up payments in the console**

Connect your own wallet at the vAPI console and follow the payment setup there. New to crypto? The console supports a smart wallet that works with a passkey or email, with no browser extension and no seed phrase. Your keys stay in your wallet, and every paid call is signed explicitly by you.

**3. Ask your agent what you need**

Search happens in the chat; payment happens with your signature in the chat.

## Route B: Via the CLI (developers)

{% hint style="info" %}
`@vapi/mcp` is not yet published to npm. The commands below build it from a checkout of the vAPI source tree, so this route currently requires access to that repository.
{% endhint %}

**1. Build and initialize the wallet**

From the repository root, with dependencies installed (`pnpm install`):

```sh
pnpm --filter @vapi/mcp build
VAPI_AGENT_CASH_HOME="$PWD/.context/agent-cash" \
  pnpm --filter @vapi/mcp exec node dist/cli.js init
```

This creates an encrypted keystore on your machine and prints a wallet address. The private key is generated locally and never leaves your machine. vAPI never holds buyer funds and never holds the keys that move them. Every payment is signed on the buyer's machine.

**2. Fund it like pocket money**

Send a small amount of Base USDC to the printed address. Keep only what your agent should be able to spend, and sweep the rest back later. Base ETH is only needed for sweep-back gas.

**3. Set spend caps**

The wallet enforces local limits before it signs anything. Defaults are $0.10 per call and $1.00 per UTC day, stored in atomic USDC units in `config.json`:

```json
{
  "spendCaps": {
    "perCallAtomic": "100000",
    "perDayAtomic": "1000000"
  }
}
```

Every quote is checked against both caps and reserved in a local ledger before signing. Any call can set a lower `maxPriceUsd` ceiling.

**4. Connect your MCP client**

```json
{
  "mcpServers": {
    "agent-cash": {
      "command": "node",
      "args": ["/absolute/path/to/dist/cli.js", "serve"],
      "env": {
        "VAPI_AGENT_CASH_HOME": "/absolute/path/to/agent-cash-home",
        "VAPI_KEYSTORE_PASSWORD": "your-keystore-passphrase"
      }
    }
  }
}
```

Restart the client after saving. Desktop MCP clients need the passphrase in the environment because they cannot show an interactive prompt. Protect this file like the wallet itself.

**5. Search, then call**

```
wallet {}
search {"query":"<capability>","kinds":["api"],"network":"eip155:8453"}
call {"id":"<search-result-ref>","network":"eip155:8453","maxPriceUsd":"0.02"}
```

The wallet tool confirms your address and balance. Search finds callable APIs. Call reads the provider's live x402 requirements, checks them against your caps and ceiling, signs one exact payment, retries the request with proof, and returns the response with the decoded settlement receipt under `payment.settlement`.

That's it. Your agent can now buy what it needs by itself, within the limits you set.


---

# 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.vapinetwork.ai/for-agents/quickstart.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.