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

# Claude Agent SDK

> Build a research agent with AgentMuxer and Claude Agent SDK.

In this guide, you'll build an agent that researches the latest TypeScript release and reports what it found, its sources, and the tool cost.

<Accordion title="Add to an existing agent">
  Register AgentMuxer in your existing `query()` options:

  ```typescript theme={null}
  import { agentmuxer } from "@agentmuxer/sdk/claude-agent";

  const options = {
    mcpServers: { agentmuxer: agentmuxer() },
    allowedTools: ["mcp__agentmuxer__*"],
  };
  ```

  `allowedTools` lets Claude call AgentMuxer tools without a separate Claude permission prompt.
</Accordion>

## Before you start

* Node.js 22.18 or later with ESM; see [compatibility](/reference/compatibility). Your host must also support the [Claude Agent SDK process requirements](https://code.claude.com/docs/en/agent-sdk/hosting).
* An [AgentMuxer application key](/sdk/applications) and credits for paid calls.
* An Anthropic API key for the model.

<Note>
  This example asks the agent to spend up to \$0.10 on one tool call. Model charges are separate. Set an [application spending limit](/sdk/applications) if you need a hard cap; the prompt alone doesn't enforce one.

  This example cannot collect payment confirmation. [Configure your account for unattended purchases](/guides/billing#unattended-applications) before running it. This setting also affects your personal agents.
</Note>

<Steps>
  <Step title="Install">
    In a new directory, create an ESM project and install the integration:

    ```bash theme={null}
    mkdir my-agent
    cd my-agent
    npm init -y
    npm pkg set type=module
    npm install @agentmuxer/sdk @anthropic-ai/claude-agent-sdk
    ```
  </Step>

  <Step title="Set your keys">
    Set both keys in your terminal:

    ```bash theme={null}
    export AGENTMUXER_API_KEY='your-application-key'
    export ANTHROPIC_API_KEY='your-model-provider-key'
    ```

    No browser sign-in is needed. Keep both keys on your server and out of source control.
  </Step>

  <Step title="Run Claude with AgentMuxer">
    Save this as `agent.ts`. The highlighted lines connect AgentMuxer and let Claude use its tools:

    ```typescript agent.ts highlight={11-12} theme={null}
    import { agentmuxer } from "@agentmuxer/sdk/claude-agent";
    import { query } from "@anthropic-ai/claude-agent-sdk";

    const prompt = `Use AgentMuxer to research the latest TypeScript release.
    Summarize three changes with sources and report the tool cost.
    Spend at most $0.10 on one tool call. Stop if it costs more.`;

    for await (const message of query({
      prompt,
      options: {
        mcpServers: { agentmuxer: agentmuxer() },
        allowedTools: ["mcp__agentmuxer__*"],
      },
    })) {
      if (message.type !== "result") continue;
      if (message.subtype !== "success") {
        throw new Error(message.errors.join("\n") || message.subtype);
      }
      if (message.is_error) {
        throw new Error(message.result);
      }
      console.log(message.result);
    }
    ```

    `query()` runs Claude's agent loop. The `for await` loop reads its messages; this example prints the final answer or throws if the run fails.
  </Step>

  <Step title="Run">
    ```bash theme={null}
    node agent.ts
    ```

    The example asks the agent to print:

    * Three TypeScript release changes.
    * Sources for the summary.
    * The actual tool cost, or the reason it could not complete the purchase.

    To try another task, edit `prompt`. Include the spending you allow.
  </Step>
</Steps>

## Customize permissions

`allowedTools` preapproves AgentMuxer calls in Claude. It does not bypass AgentMuxer payment confirmation or hide unlisted tools. Configure `disallowedTools` or `canUseTool` on your Claude agent when you need more control.

The helper's `tools` option sets per-tool policies, not a tool allowlist. It also accepts native `timeout` and `alwaysLoad` options. See the [SDK reference](/reference/sdk) for details.

<CardGroup cols={2}>
  <Card title="Applications and keys" icon="key" href="/sdk/applications">
    Manage credentials and spending limits.
  </Card>

  <Card title="Run in production" icon="server" href="/sdk/production">
    Handle deployment, failures, and tracing.
  </Card>
</CardGroup>