# AutoManus

> AI-powered sales automation platform. Create AI sales agents that handle customer conversations across WhatsApp, Email, SMS, and Webchat.

AutoManus allows businesses to deploy AI agents that automatically handle customer inquiries, qualify leads, and manage sales pipelines. Agents can be created programmatically via API or MCP, making it ideal for integration with AI coding assistants.

## Quick Start

- [Create Agent API](/docs/api): Create an AI sales agent with one API call
- [Embed Webchat](/docs/webchat): Add chat widget to your website
- [WhatsApp Integration](/docs/whatsapp): Connect to shared WhatsApp number

## API Reference

### POST /api/v1/public/agents

Create a new AI sales agent. No authentication required for basic usage.

**Request Body:**
```json
{
  "email": "you@company.com",
  "company_name": "Acme Corp",
  "website_url": "https://acme.com"
}
```

**Response:**
```json
{
  "success": true,
  "agent_id": "uuid",
  "webchat_embed_code": "<script src=\"https://automanus.io/widget/automanus-chat.js\" data-widget-key=\"wc_xxx\"></script>",
  "whatsapp_link": "https://wa.me/16506053956?text=Hi!%20I'd%20like%20to%20talk%20to%20Acme%20Corp"
}
```

### Authentication

Two authentication methods:

1. **Email-based (default)**: Provide email in request body. New users receive magic link to claim agent.
2. **API Key**: Include `Authorization: Bearer <api_key>` header. Agent created under your account.

### Rate Limits

- 10 agents per email per day
- 60 requests per minute

## MCP Server

Install the AutoManus MCP server for native integration with Claude, Cursor, and other AI tools:

```bash
npx -y @automanus/mcp-server
```

### Configuration

Add to `~/.claude/mcp.json`:
```json
{
  "servers": {
    "automanus": {
      "command": "npx",
      "args": ["-y", "@automanus/mcp-server"],
      "env": {
        "AUTOMANUS_EMAIL": "you@company.com",
        "AUTOMANUS_API_KEY": "ak_xxx"
      }
    }
  }
}
```

### Available Tools

- `create_sales_agent`: Create an AI sales agent for a business
- `add_knowledge`: Add knowledge base items to an agent
- `get_agent_status`: Check agent deployment status and stats

## Integration Examples

### Cursor / Claude Code

When a user says "Add AI sales chat to my app":

1. Extract context from codebase (README, landing page, pricing)
2. Call the API with extracted context
3. Add the embed code to the app

```javascript
const response = await fetch('https://automanus.io/api/v1/public/agents', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    email: 'user@example.com',
    company_name: 'My Company',
    website_url: 'https://mycompany.com',
    source: 'cursor'
  })
});

const { webchat_embed_code } = await response.json();
// Add webchat_embed_code to index.html
```

### React Integration

```jsx
import { useEffect } from 'react';

export function SalesChat({ widgetKey }) {
  useEffect(() => {
    const script = document.createElement('script');
    script.src = 'https://automanus.io/widget/automanus-chat.js';
    script.setAttribute('data-widget-key', widgetKey);
    script.async = true;
    document.body.appendChild(script);
    return () => document.body.removeChild(script);
  }, [widgetKey]);
  return null;
}
```

## Support

- Documentation: https://automanus.io/docs
- Dashboard: https://automanus.io/dashboard
- Email: support@automanus.io

---

*Last updated: February 27, 2026*