Dragon + LangGraph Examples

A guided path showing what Dragon adds to a LangGraph agent graph on an HPC cluster — distributed agent execution, node placement, on-cluster LLM inference, node-local data locality, and real-time supervision of live experiments. Every example demonstrates a capability that plain LangGraph does not have. Read them in order; each adds one new Dragon ability.

These examples run under the Dragon runtime, not plain python:

dragon examples/dragon_ai/langgraph/01_quickstart.py

For the internals, see LangGraph Integration.

• 01 — Quickstart
• 02 — Multinode Placement
• 03 — On-Cluster Inference
• 04 — Real-Time Supervision
• 05 — Data-Locality Crossover
• 06 — Self-Driving Lab