Skip to content

langgraph-jev

langgraph-jev is an open-source LangGraph/LangChain integration for Jev, TypeSafe's typed probabilistic decision API. JevNode wraps it as a decision node you drop into an agent graph -- it's not an LLM and not an LLM wrapper.

Jev evaluates typed questions against state and returns typed answers with probabilities and confidence attached. There's no text generation and nothing to parse -- your code branches on, sorts by, and routes with the result directly.

For example, a work_ticket graph might run an "understand" step over a customer request, hand the structured output to a JevNode for typed decisions (work type, priority, whether it needs an engineer), and then route to a coding, support, or docs worker based on those decisions. See the examples page for the full version of this.

Terminal output of examples/basic.py

Where to go next

  • Quick start -- install, set your API key, define your first question.
  • API reference -- the full public surface: JevClient, JevNode, JevRunnable, question builders, errors, and result types.
  • Examples -- runnable scripts covering LangGraph nodes, confidence thresholds, and the WorkTicket pattern.

Source

The project lives at github.com/iroy2000/langgraph-jev. Issues and pull requests are welcome -- see CONTRIBUTING.md in the repo.