react_agent

ReAct Agent (Graph API)

Demonstrates the most common LangGraph pattern: a tool-calling agent that loops between deciding and acting, using conditional edges for routing.

What This Sample Demonstrates

• Defining a StateGraph with an agent->tools loop
• Using add_conditional_edges for conditional routing (call tool or finish)
• Accumulating conversation history with Annotated[list, operator.add]
• The full ReAct cycle: think -> act -> observe -> repeat

How It Works

1. The agent node examines the conversation history and decides the next action.
2. If a tool is needed, should_continue routes to the tools node.
3. The tools node executes the tool and appends the result to history.
4. Control returns to agent, which decides again — loop or finish.
5. When the agent has enough information, should_continue routes to END.

START -> agent -> tools -> agent -> tools -> agent -> END

Running the Sample

Prerequisites: uv sync --group langgraph and a running Temporal dev server (temporal server start-dev).

# Terminal 1
uv run langgraph_plugin/graph_api/react_agent/run_worker.py

# Terminal 2
uv run langgraph_plugin/graph_api/react_agent/run_workflow.py

Files

File	Description
workflow.py	AgentState, node functions, should_continue router, graph definition, and ReactAgentWorkflow
run_worker.py	Builds graph, registers with LangGraphPlugin, starts worker
run_workflow.py	Executes the agent workflow and prints the answer