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
/* Copyright 2026 Marimo. All rights reserved. */
import { useAtomValue } from "jotai";
import { createReducerAndAtoms } from "@/utils/createReducer";
import { generateUUID } from "@/utils/uuid";
export interface TerminalCommand {
id: string;
text: string;
timestamp: number;
}
export interface TerminalState {
pendingCommands: TerminalCommand[];
isReady: boolean;
}
function initialState(): TerminalState {
return {
pendingCommands: [],
isReady: false,
};
}
const {
reducer,
createActions,
valueAtom: terminalStateAtom,
useActions,
} = createReducerAndAtoms(initialState, {
addCommand: (state, text: string) => {
const command: TerminalCommand = {
id: generateUUID(),
text,
timestamp: Date.now(),
};
return {
...state,
pendingCommands: [...state.pendingCommands, command],
};
},
removeCommand: (state, commandId: string) => ({
...state,
pendingCommands: state.pendingCommands.filter(
(cmd) => cmd.id !== commandId,
),
}),
setReady: (state, isReady: boolean) => ({
...state,
isReady,
}),
clearCommands: (state) => ({
...state,
pendingCommands: [],
}),
});
/**
* React hook to get the terminal state.
*/
export const useTerminalState = () => useAtomValue(terminalStateAtom);
/**
* React hook to get the terminal actions.
*/
export function useTerminalActions() {
return useActions();
}
export const exportedForTesting = {
reducer,
createActions,
initialState,
};