Automate browser based workflows with AI
Python aiapiautomationbrowserbrowser-automationcomputergptllmplaywrightpowerautomatepuppeteerpythonrpaseleniumvisionworkflow
import { create } from "zustand";
import { type BlockSearchTarget } from "@/routes/workflows/studio/blockSearch";
type WorkflowBlockSearchHandle = {
// Searchable blocks in canvas order, read fresh on each popover open.
getTargets: () => Array<BlockSearchTarget>;
// Selects the block and centers the canvas on it.
focusBlock: (nodeId: string) => void;
};
type WorkflowBlockSearchState = {
// Registered by the studio's embedded FlowRenderer (which owns the React
// Flow instance) so the Editor pane header's search can jump the canvas
// without owning it. Null while no searchable canvas is mounted — the
// header hides the control.
handle: WorkflowBlockSearchHandle | null;
registerHandle: (handle: WorkflowBlockSearchHandle | null) => void;
};
export const useWorkflowBlockSearchStore = create<WorkflowBlockSearchState>(
(set) => ({
handle: null,
registerHandle: (handle) => set({ handle }),
}),
);