[Weave] How to associate a trace with a specific run

weave/guides/tools/weave-in-workspaces.mdx

@@ -87,6 +87,35 @@ To navigate to a workspace from the UI:
 
 For more information about workspaces, see [View experiments results](https://docs.wandb.ai/models/track/workspaces).
 
+## Associate traces with a specific W&B run
+
+By default, Weave automatically detects and associates traces with the active `wandb.run`. If you need to associate traces with a specific run that is not the global `wandb.run`, use `set_wandb_run_context`. You can associate traces with a specific run and run step by defining values for the method's `run_id` and `step` arguments:
+
+```python lines highlight={6}
+import weave
+
+client = weave.init("<your-team-name>/<your-project-name>")
+
+# Associate all subsequent traces with a specific run.
+client.set_wandb_run_context(run_id="<your-run-ID>", step=<integer>)
+
+@weave.op
+def my_function(x):
+    return x * 2
+
+my_function(10)  # This trace is linked to "my-run-id" at step 5.
+```
+
+The `run_id` is the run's short ID (for example, `38m4t5ja`), not the full `entity/project/run_id` path. You can find a run's short ID in the run URL (`https://wandb.ai/team/project/runs/<run_id>`) or by [accessing `wandb.run.id` in code](/models/runs/run-identifiers#find-and-customize-a-runs-id-or-name).
+
+To stop the override and fall back to the global `wandb.run`, call `clear_wandb_run_context`:
+
+```python lines
+client.clear_wandb_run_context()
+
+my_function(20)  # This trace uses the global wandb.run (if any).
+```
+
 ## Reference W&B artifacts in a Weave trace
 
 You can reference [W&B artifacts](/models/artifacts) (such as models, datasets, and checkpoints) in your Weave traces. This creates a clickable link in the Weave UI that navigates directly to the artifact's details page, helping you track which artifact versions were used during specific operations.