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

# Introduction

> Deploy long-running AI agents from a single config file.

**swarmlord** runs AI agents that you define in one `swarmlord.jsonc` file.

```jsonc swarmlord.jsonc theme={null}
{
  "name": "hello-agent",
  "model": { "providerId": "openrouter", "modelId": "anthropic/claude-opus-4.8" },
  "instructions": "You are a helpful agent."
}
```

```bash theme={null}
swarmlord deploy && swarmlord run -m "Say hi."
```

That's the whole thing. The agent is now reachable from `curl`, the SDK, the CLI, and the dashboard.

## What you get

* **Persistent sessions** — multi-turn memory, fork, abort, and resume from any disconnect.
* **Built-in web tools** — `websearch` and `webfetch` ready for every agent.
* **MCPs in one line** — Sendblue, GitHub, Slack, Playwright, more.
* **Schedule and trigger** — cron expressions and webhook URLs come with the bundle.
* **Real billing** — every model call recorded in USD.

## Where to next

<CardGroup cols={2}>
  <Card title="Quickstart" icon="rocket" href="/quickstart">
    Deploy your first agent in 5 minutes.
  </Card>

  <Card title="Concepts" icon="diagram-project" href="/concepts/agents">
    Agents, sessions, MCPs.
  </Card>

  <Card title="CLI" icon="terminal" href="/guides/cli">
    Every command, in depth.
  </Card>

  <Card title="SDK" icon="code" href="/guides/sdk">
    The TypeScript client.
  </Card>
</CardGroup>