What if I told you that you could go from a simple English description like “Create a website for uploading and storing client PDFs, with CRUD, and have a search interface that allows you to talk to those documents and stores all conversations in a sql database” to a fully deployed, production-ready application in under an hour? Not a cute laptop demo, but globally available fully scaleable and secure app? In under an hour? Without writing any infrastructure code, configuring servers, or wrestling with deployment pipelines?
With this AI tool, the capabilities to streamline the entire process from idea to production are at your fingertips, automating key functions such as infrastructure setup, deployment, and workflow management.
This isn’t science fiction—it’s the reality of combining Cursor’s Agent Mode with LiquidMetal’s Raindrop Platform-as-a-Service (PaaS). We’ve created a set of intelligent rules that enable a range of functions and capabilities within the AI tool, teaching Cursor how to architect, build, and deploy cloud-native applications using natural language commands.
Traditional AI coding assistants require constant hand-holding. You specify context, guide every step, and manually execute commands. Cursor’s Agent Mode changes everything:
Normal Mode limitations:
Agent Mode superpowers:
When you combine Agent Mode with LiquidMetal’s specialized rules, you get an AI that doesn’t just write code—it architects, builds, and deploys complete applications.
Getting started takes less than 5 minutes:
Rules can be stored in a .cursor/rules folder, which is still supported but may be phased out in favor of newer formats for better flexibility and control. Each rule file is written in the mdc format to ensure compatibility and ease of use.
Our rules teach Cursor everything it needs to know about LiquidMetal’s platform:
Here’s where it gets exciting. With our rules in place, your development workflow becomes incredibly simple:
You: “Create a Raindrop application called customer-feedback-analyzer that processes customer reviews, stores them in a database, and sends summaries to a queue for further processing.”
This user request acts as a query, which the system interprets to generate code and configure the workflow accordingly. Users can issue natural language queries to guide the AI in generating code and managing workflows.
Cursor Agent Mode:
No infrastructure knowledge required. No DevOps complexity. Just pure focus on your business logic.
The versatility is remarkable. Our rules enable Cursor to architect virtually any type of application:
Each application leverages LiquidMetal’s building blocks — services, observers, actors, buckets, SmartBuckets, and SQL databases — to assemble intelligently based on your requirements.
Traditional cloud application development follows a painful timeline:
With Cursor Agent Mode + LiquidMetal:
By automating simple tasks and routine coding activities, the system helps streamline software development and significantly boosts productivity. This allows developers to focus on higher-value work while AI accelerates the entire workflow.
We’re not exaggerating. The Raindrop PaaS handles all infrastructure complexity, while Cursor handles all the architectural decisions and code generation.
Our Cursor rules aren’t just simple templates—they’re a comprehensive knowledge system that teaches Cursor to think like an expert cloud architect on LiquidMetal AI:
Intelligent Resource Management: Cursor learns when to use SmartBuckets vs. regular buckets, when to implement queues for processing, and how to structure SQL schemas for optimal performance.
Security by Default: Every generated application includes proper request validation, security headers, authentication handling, and input sanitization—no security afterthoughts.
Production-Ready Code: Generated applications include error boundaries, meaningful error messages, appropriate logging, and graceful edge case handling. The system helps developers fix bugs and integrate new features, ensuring that each addition works seamlessly with the existing code and the rest of the project. Developers benefit from automated bug detection and feature integration, which streamlines the process of maintaining high-quality, production-ready software.
TypeScript Excellence: Modern ES modules, proper type definitions, and clean, maintainable code structure that follows current best practices.
This integration represents something bigger than just faster development — it’s about democratizing access to sophisticated cloud applications.
Agentic AI systems and large language models drive innovation by supporting decision making and problem solving across projects. These systems process vast amounts of data, track user preferences, and enable AI agents to operate autonomously within each project. AI agents interact with external tools and use user rules, cursor rules, and project-specific files and folders to organize, format, and process data efficiently. For example, documentation and support tools provide clear guidance for creating, managing, and organizing projects, making it easier for users to leverage the full capabilities of agentic AI systems. Features are enabled through specific settings, and proper formatting and organization of files and folders are essential for maintaining clarity and consistency in large projects.
Previously, building production-ready cloud apps required:
Now, domain experts can focus entirely on their business logic while AI handles the technical architecture. A marketing professional can build a customer sentiment analyzer. A researcher can create a data processing pipeline. An entrepreneur can prototype their MVP.
A marketing professional can build a customer sentiment analyzer. A researcher can create a data processing pipeline. An entrepreneur can prototype their next big idea—and because Raindrop handles all scaling automatically, their “prototype” is actually production-ready from day one.
The barrier between having an idea and having a deployed application has virtually disappeared.
Ready to experience this yourself? Here’s your quickstart guide:
We’re witnessing a fundamental shift in how applications are built. The combination of intelligent AI agents with platform-specific knowledge creates possibilities that seemed impossible just months ago.
This is just the beginning. As we add more building blocks to LiquidMetal and refine our Cursor rules, the scope of what’s possible will only expand. We’re moving toward a world where the speed of thought becomes the speed of development.
The question isn’t whether AI will transform software development—it’s whether you’ll be part of the transformation or left behind by it.
Ready to build the future? Start with our Cursor integration and see how quickly you can turn your next idea into reality.
Try the integration yourself and share your creations with us. We’re building a community of developers who are pushing the boundaries of AI-assisted development, and we’d love to see what you create.
Join the conversation on our Discord
Get started: go to our docs