JupyterLab — langstage-jupyter

The JupyterLab stage: a chat sidebar that gives your LangGraph agent access to notebooks and files, for natural-language data-science work without leaving the lab.

dkedar7/langstage-jupyter PyPI

Quickstart

pip install langstage-jupyter

# Instead of `jupyter lab`, use the launcher (auto-configures the server):
langstage-jupyter                            # default agent (needs ANTHROPIC_API_KEY)
langstage-jupyter --demo                      # keyless echo agent
langstage-jupyter -a my_agent.py:graph        # your agent
langstage-jupyter --show-config               # resolved config + sources

The launcher auto-detects a port, generates an auth token, sets the required environment, and starts JupyterLab with the chat sidebar wired up.

The chat sidebar, light and dark — it follows JupyterLab's own theme.

Highlights

• Notebook tools — create, edit, and execute cells from chat.
• Context-aware — sends the focused widget, current directory, and selection to your agent.
• Human-in-the-loop review of agent actions.
• Uses the same agent spec and config as every other stage; pick the agent with -a/--agent or LANGSTAGE_AGENT_SPEC.

Screenshots show small timestamp masks

The sidebar images above come from the visual-regression baselines, which mask volatile timestamps. Functionally representative; cleaner marketing shots are a planned follow-up.