A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.
Python artificial-intelligencedagdata-sciencedata-visualizationdataflowdeveloper-toolsmachine-learningnotebookspipelinepythonreactivesqlweb-app
docs: autorun agents (#7217)
Myles Scolnick11d132a6e3f6
diff --git a/docs/guides/editor_features/agents.md b/docs/guides/editor_features/agents.mdindex 3bb4e1eb57..b5ce09d3f4 100644--- a/docs/guides/editor_features/agents.md+++ b/docs/guides/editor_features/agents.md@@ -73,6 +73,17 @@ See login and authentication instructions in the [Gemini CLI documentation](httpIf you have terminal access enabled in marimo, you can run agent connection commands directly from the agent panel using the terminal button.+!!! tip "Auto-run on agent edits"++ By default, when an agent modifies your notebook, cells are marked as stale instead of running automatically. To have cells run automatically when the agent saves changes, add this configuration to your `pyproject.toml`:++ ```toml+ [tool.marimo.runtime]+ watcher_on_save = "autorun"+ ```++ This provides a more seamless experience when working with agents, as you'll see results immediately after the agent makes changes.+## Custom agents!!! info "Custom agents"