basic

Basic LLM Workflow

A one-shot workflow that sends a prompt to OpenAI and returns the response. A very simple example of LangSmith tracing with Temporal.

See the parent README for prerequisites.

Running

# Terminal 1 — start the worker
uv run --group langsmith-tracing python -m langsmith_tracing.basic.worker

# Terminal 2 — run the workflow
uv run --group langsmith-tracing python -m langsmith_tracing.basic.starter

Trace Structure

add_temporal_runs=False (default)

Only @traceable and wrap_openai spans appear. The client-side @traceable is the root, and workflow/activity traces nest under it via context propagation.

Basic LLM Request                    (@traceable, client-side)
└── Ask: What is Temporal?           (@traceable, workflow)
    └── Call OpenAI                   (@traceable, activity)
        └── ChatOpenAI                (automatic via wrap_openai)

add_temporal_runs=True

Pass --add-temporal-runs to both the worker and starter:

uv run --group langsmith-tracing python -m langsmith_tracing.basic.worker --add-temporal-runs
uv run --group langsmith-tracing python -m langsmith_tracing.basic.starter --add-temporal-runs

Temporal operation spans are added. StartWorkflow/RunWorkflow and StartActivity/RunActivity appear as sibling pairs:

Basic LLM Request                       (@traceable, client-side)
├── StartWorkflow:BasicLLMWorkflow      (automatic, Temporal plugin)
└── RunWorkflow:BasicLLMWorkflow        (automatic, Temporal plugin)
    └── Ask: What is Temporal?          (@traceable, workflow)
        ├── StartActivity:call_openai   (automatic, Temporal plugin)
        └── RunActivity:call_openai     (automatic, Temporal plugin)
            └── Call OpenAI             (@traceable, activity)
                └── ChatOpenAI          (automatic via wrap_openai)