llms.txt Content
# Liquid Metal Documentation
> Liquid Metal is a cloud development platform that enables developers to build and deploy production-ready AI applications in seconds.
> The platform provides a streamlined experience for creating, managing, and deploying cloud-native applications with zero infrastructure management.
This documentation website provides comprehensive information about the Liquid Metal platform, including guides, SDK documentation, and reference materials.
## Core Documentation
- [Actors](/raw-markdown/concepts/actors.md): Stateful compute units that maintain persistent data and handle requests with unique identity.
- [AI Models](/raw-markdown/concepts/ai-models.md): Overview of AI model types available in Raindrop for text generation, image processing, speech recognition, and more.
- [Annotations](/raw-markdown/concepts/annotations.md): Metadata management system for organizing configuration, documentation, and application context.
- [Bucket Storage](/raw-markdown/concepts/bucket-storage.md): Object storage for files and data that scales from simple documents to massive media libraries.
- [Claude Code + Raindrop MCP](/raw-markdown/concepts/claude-code-raindrop-mcp.md): How the Raindrop MCP server transforms Claude Code into a complete application development platform for building infrastructure and deploying real applications.
- [Key-Value Storage](/raw-markdown/concepts/kv.md): Fast, simple data storage for caching, sessions, and temporary data persistence in Raindrop applications.
- [Observers](/raw-markdown/concepts/observers.md): Event-driven components that react to changes in buckets and queues, enabling reactive architecture patterns.
- [Queues](/raw-markdown/concepts/queues.md): Asynchronous message passing and decoupled system design for background processing and workflow coordination.
- [Guided Development Flow](/raw-markdown/concepts/raindrop-development/guided-development.md): How the structured workflow prevents chaos and ensures complet