Rivet

Rivet is a visual programming environment for building AI agents that use large language models. Instead of writing complex chains of prompts in code, you can design them as visual graphs — connecting nodes, setting up logic, and seeing how data flows between steps. It's built by Ironclad, the digital contracting company, and they use it themselves to power their own AI features. The tool is open-source and free to use.

What makes Rivet stand out is its focus on the whole development lifecycle, not just prototyping. You can build a prompt graph visually, then run it directly inside your own application using Rivet's remote executor. The debugger lets you watch the execution of your prompt chains in real-time, which is incredibly useful when you're trying to figure out why an AI agent made a weird decision. Since Rivet graphs are just YAML files, you can version them in Git and review changes with your team using standard code review tools. That's a big deal for teams that need to collaborate on complex AI logic.

Rivet is best for developers and teams building production AI applications — think chatbots, virtual assistants, or any tool that chains multiple LLM calls together. If you're tired of debugging prompt chains by reading through logs, or you want to make it easier for non-engineers to understand and contribute to AI workflows, Rivet is worth a look. It's also a great fit for teams that need to iterate quickly on agentic logic and want a clear visual representation of what the AI is actually doing.