CREator
CREator: An AI-powered visual builder to design, deploy, and simulate Chainlink CRE workflows.
What it is
CREator is a web-based, AI-enhanced visual development environment exclusively tailored for the new Chainlink Runtime Environment (CRE). It allows developers and non-technical users alike to build complex, decentralized automated workflows by just dragging and dropping nodes on a canvas.
How it works: Users can manually assemble workflows using predefined nodes representing Triggers, Data Sources, Logic, Chainlink Services (Oracles, CCIP, Functions), and Smart Contracts. Alternatively, users can prompt an integrated AI Web3 Assistant to automatically generate the workflow infrastructure from natural language. From the same UI, users can review, compile, and deploy Solidity smart contracts directly to testnets using MetaMask, without context-switching to Remix. Finally, clicking "Prove" instantly auto-generates the underlying CRE syntax (main.ts and workflow.yaml) and executes the Chainlink CRE simulator on the backend, streaming real-time execution logs back to the user.
The problem it solves: Writing CRE workflows from scratch has a high barrier to entry due to stringent architectural requirements and multiple moving parts (wallets, TS scripts, YAML configs, smart contracts). CREator acts as a unifying orchestration layer that abstracts away the boilerplate, validates the logic against Chainlink rules, and transforms hours of setup into minutes of visual configuration.
How it Works
- Frontend: Built with React, TypeScript, and Vite. We heavily utilized
reactflowto create the interactive node-based canvas. For Web3 integrations, we usedethers.js,wagmi, andRainbowKitto handle wallet connections and smart contract deployments. We even built an in-browser Solidity compilation step (solc). - Backend: An Express.js (Node.js) server that handles the orchestration. It manages the communication with the DeepSeek LLM for the AI assistant and acts as a bridge to execute the Chainlink CRE CLI natively, returning standard output logs to the simulation modal.
- Architecture: The system takes a JSON representation of the visual canvas, validates it, and dynamically translates it into production-ready
main.tsscripts andworkflow.yamlconfiguration profiles.
Links
Created by
- Franco Ayala
- jhamil mamani