# Lotus - Browser Automation Skills
Create browser automation skills by demonstration. Just show Lotus what you want to automate, and it creates a reusable skill.
## Features
- **Record by demonstration** - Open a browser, perform actions, narrate what you're doing
- **Voice narration** - Speak to describe your workflow while recording
- **AI-powered refinement** - Lotus understands your intent and creates robust automation
- **Reusable skills** - Call your skills anytime from Cursor chat
## Quick Start
1. **Get an API key** at [uselotus.ai](https://uselotus.ai)
2. **Set your API key** in Cursor:
- Open Settings (`Cmd+,` / `Ctrl+,`)
- Search "Lotus"
- Enter your API key
3. **Create a skill**:
- Open Cursor chat (`Cmd+L` / `Ctrl+L`)
- Ask: "Create a skill to check my calendar"
- A browser opens inside Cursor
- Demonstrate the workflow while speaking
- Say "I'm done" when finished
4. **Use your skill**:
- Ask Cursor to run your skill by name
- Or use the tool directly: `lotus_check_my_calendar`
## How It Works
1. **Recording** - Lotus opens a cloud browser where you demonstrate your workflow
2. **Narration** - Your voice is captured to understand context and intent
3. **Refinement** - AI analyzes your actions and creates a robust automation script
4. **Execution** - Run your skill anytime with natural language or direct tool calls
## Commands
- `Lotus: Open Browser` - Manually open the Lotus browser panel
## Settings
| Setting | Description |
|---------|-------------|
| `lotus.apiKey` | Your Lotus API key (get one at [uselotus.ai](https://uselotus.ai)) |
## Requirements
- Cursor IDE
- Lotus API key
## Support
- [Documentation](https://uselotus.ai/docs)
- [GitHub Issues](https://github.com/chrisvin-jabamani/lotus-mcp/issues)
- [Website](https://uselotus.ai)
## License
MIT
