langsmith_tracing

LangSmith Tracing (Functional API)

Demonstrates combining LangGraphPlugin (durable task execution) with Temporal's LangSmithPlugin for full tracing of LLM calls through Temporal workflows, using LangGraph's @task and @entrypoint decorators.

What This Sample Demonstrates

• Using LangSmithPlugin on the Temporal client for automatic trace propagation
• Using LangGraphPlugin on the Worker for durable LangGraph execution
• @traceable in three places: on the @task (Activity) itself, on a helper called from inside the @task, and on a helper called from inside the @entrypoint (Workflow)
• Both plugins working together: durability + observability

How It Works

1. The Temporal client is created with LangSmithPlugin(add_temporal_runs=True).
2. A Worker registers the chat task with LangGraphPlugin.
3. When the Workflow runs, the chat task executes as a Temporal Activity.
4. @traceable decorators emit trace data to LangSmith for the task, an in-task helper, and an in-entrypoint helper.

Running the Sample

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

export ANTHROPIC_API_KEY='your-key'
export LANGCHAIN_API_KEY='your-key'

uv run langgraph_plugin/functional_api/langsmith_tracing/main.py

Traces will appear in your LangSmith dashboard.

Files

File	Description
workflow.py	@traceable chat task + helpers, @entrypoint, and ChatFunctionalWorkflow
main.py	Starts a Worker and executes the Workflow in a single process