Rivet is an open-source visual programming environment for building AI agents with large language models (LLMs). It allows teams to design, debug, and collaborate on complex LLM prompt graphs visually, then run them directly in their applications. Rivet graphs are stored as YAML files, enabling version control and code review through standard tools. The environment includes features for visualizing and building complex chains for production use, remote debugging to observe prompt chain execution in real-time, and collaboration via repository-based graph sharing. Rivet is built and used by Ironclad Research, the team behind Ironclad's digital contracting platform and Ironclad Contract AI (CAI), a virtual contract assistant developed with Rivet.
Key Benefits
- Visual environment for designing and debugging LLM prompt graphs
- Remote debugger for observing execution in real-time
- Graphs are YAML files, enabling version control and code review
- Open-source and free to use