Automate browser based workflows with AI
Python aiapiautomationbrowserbrowser-automationcomputergptllmplaywrightpowerautomatepuppeteerpythonrpaseleniumvisionworkflow
import { create } from "zustand";
type RunViewState = {
// The frame the user is inspecting. null means "follow the live edge" while
// running, or the final frame once finalized.
pinnedFrameId: string | null;
// Bumped on every pinFrame so the Browser pane can react to a re-pin of the
// already-selected frame (the id alone wouldn't change).
pinNonce: number;
// Loop-iteration scope of a pinned container block. Not carried in ?active=,
// so the Browser pane reads it here to resolve the iteration's screenshot.
activeIteration: number | null;
pinFrame: (id: string, iteration?: number | null) => void;
jumpToLive: () => void;
reset: () => void;
};
export const useRunViewStore = create<RunViewState>((set) => ({
pinnedFrameId: null,
pinNonce: 0,
activeIteration: null,
pinFrame: (id, iteration = null) =>
set((state) => ({
pinnedFrameId: id,
activeIteration: iteration,
pinNonce: state.pinNonce + 1,
})),
jumpToLive: () => set({ pinnedFrameId: null, activeIteration: null }),
reset: () => set({ pinnedFrameId: null, activeIteration: null }),
}));