Interactive components

Create approval flows where the agent pauses and waits for human input.

Gen UI Interrupt (Frontend Tool)

Waking up the demo

What is this?

Interactive generative UI creates flows where the agent pauses execution and waits for user input before continuing. This enables approval workflows, confirmation dialogs, and any scenario where human judgment is needed mid-execution.

When should I use this?

Use interactive generative UI when you need:

  • Approval/rejection flows (e.g. "Run this command?")
  • User decisions that the agent should know about
  • Confirmation dialogs with structured responses
  • Any flow where the agent pauses for human judgment

How it works in code

useHumanInTheLoop: nothing to wire on the agent

The AG-UI Mastra adapter forwards each tool you register with useHumanInTheLoop to the agent as a client tool. When the model calls it, the run hands control to the browser, and the user's answer comes back as the tool result. Do not declare the tool on the Mastra agent.

useInterrupt: return suspend() from a tool

A Mastra tool pauses with suspend(), not interrupt(). Declare a suspendSchema for the payload your UI reads and a resumeSchema for the answer it sends back. The @ag-ui/mastra bridge sends the pause to the frontend as an AG-UI interrupt. When the user answers, Mastra runs execute again with the answer in resumeData.

src/mastra/tools/interrupt.ts — schedule_meeting suspend tool
// Native interrupt backend for the `gen-ui-interrupt` + `interrupt-headless`
// cells (OSS-383). Unlike the old Strategy-B workaround (a frontend-only
// `schedule_meeting` satisfied by `useHumanInTheLoop`), this is a REAL Mastra
// suspend tool: the model calls `schedule_meeting`, the tool `suspend()`s with
// a time-picker payload, and the @ag-ui/mastra bridge maps that to an AG-UI
// interrupt (legacy `on_interrupt` CUSTOM event + the standard `RUN_FINISHED`
// interrupt-outcome, on by default in the v1 bridge). The frontend `useInterrupt`
// hook renders the picker and `resolve(...)`s, which resumes the run — the tool's
// `execute` is re-invoked with `resumeData` carrying the user's selection.
//
// Load-bearing (see the Mastra capability-map memory):
//   - `return suspend(...)` DIRECTLY. `await suspend(); return x` completes the
//     tool, so under fast streaming the agentic loop continues past the pause.
//   - Resume needs instance `storage` (configured on `new Mastra({ storage })`)
//     and the suspend-chunk runId — both handled by the bridge + src/mastra/index.ts.
import { createTool } from "@mastra/core/tools";
import { z } from "zod";

export interface InterruptTimeSlot {
  label: string;
  iso: string;
}

// Mirror of `src/app/demos/_shared/interrupt-fallback-slots.ts` so the backend
// supplies future-relative candidate slots inside the suspend payload (the
// frontend only falls back to its own generator if these are absent).
function atLocal(date: Date, hour: number, minute = 0): Date {
  return new Date(
    date.getFullYear(),
    date.getMonth(),
    date.getDate(),
    hour,
    minute,
    0,
    0,
  );
}

function nextMonday(from: Date): Date {
  const day = from.getDay();
  let offset = (1 - day + 7) % 7;
  if (offset <= 1) offset += 7;
  const next = new Date(from);
  next.setDate(from.getDate() + offset);
  return next;
}

export function generateCandidateSlots(
  now: Date = new Date(),
): InterruptTimeSlot[] {
  const tomorrow = new Date(now);
  tomorrow.setDate(now.getDate() + 1);
  const monday = nextMonday(now);
  const candidates: Array<[string, Date]> = [
    ["Tomorrow 10:00 AM", atLocal(tomorrow, 10)],
    ["Tomorrow 2:00 PM", atLocal(tomorrow, 14)],
    ["Monday 9:00 AM", atLocal(monday, 9)],
    ["Monday 3:30 PM", atLocal(monday, 15, 30)],
  ];
  return candidates.map(([label, date]) => ({
    label,
    iso: date.toISOString(),
  }));
}

export const scheduleMeetingInterruptTool = createTool({
  id: "schedule_meeting",
  description:
    "Ask the user to pick a meeting time. Surfaces an interactive time-picker " +
    "to the user and returns their selection. Call this whenever the user asks " +
    "to book or schedule a meeting.",
  inputSchema: z.object({
    topic: z
      .string()
      .describe("What the meeting is about (e.g. 'Intro with sales')."),
    attendee: z
      .string()
      .optional()
      .describe("Who the meeting is with (e.g. 'Alice'), if known."),
  }),
  suspendSchema: z.object({
    topic: z.string(),
    attendee: z.string().optional(),
    slots: z.array(z.object({ label: z.string(), iso: z.string() })),
  }),
  resumeSchema: z.object({
    chosen_time: z.string().optional(),
    chosen_label: z.string().optional(),
    cancelled: z.boolean().optional(),
  }),
  // Mastra createTool execute signature is `(inputData, executionContext)`:
  // the validated INPUT is the first arg; the ToolExecutionContext is the
  // second. `suspend` / `resumeData` live under `executionContext.agent`
  // (the `AgentToolExecutionContext` sub-object), NOT at the top level.
  // Destructuring them off `executionContext` directly yields `undefined`, so
  // `return suspend(...)` throws `suspend is not a function`; the model gets a
  // tool-error, re-calls, and the loop spins to the step cap with no
  // `tool-call-suspended` chunk. Mirrors the proven @ag-ui/mastra dojo tool.
  execute: async (inputData, executionContext) => {
    const { suspend, resumeData } = executionContext?.agent ?? {};
    // Second pass: the user resolved the interrupt — the run resumes here with
    // their selection. Return a short confirmation the agent narrates.
    if (resumeData) {
      if (resumeData.cancelled) {
        return "The user cancelled — no meeting was scheduled.";
      }
      const when =
        resumeData.chosen_label ?? resumeData.chosen_time ?? "the chosen time";
      return `Scheduled "${inputData.topic}" for ${when}.`;
    }
    // First pass: suspend with the picker payload. Returned directly so the
    // agentic loop pauses here instead of continuing.
    return suspend?.({
      topic: inputData.topic,
      attendee: inputData.attendee,
      slots: generateCandidateSlots(),
    });
  },
});

Two details fail without an error:

  • suspend and resumeData are on executionContext.agent, the second argument of execute. Read from the top level of the context, they are undefined, and the model retries the tool until the run reaches its step limit.
  • return suspend(...) directly. With await suspend() followed by another return, the tool completes and the agent keeps going past the pause.

Give the Mastra instance storage

Resume reloads the paused run from storage, so the Mastra instance that registers the agent needs storage:

src/mastra/index.ts
import { Mastra } from "@mastra/core/mastra";
import { LibSQLStore } from "@mastra/libsql";

export const mastra = new Mastra({
  agents: { interruptAgent },
  storage: new LibSQLStore({
    id: "mastra-storage",
    url: ":memory:",
  }),
});

Read the payload under suspendPayload

The bridge wraps your suspend() value, so useInterrupt does not receive it as is:

  • In the custom-event flow, event.value is a JSON string. Parse it and read suspendPayload.
  • In the standard AG-UI flow, the payload is in interrupt.metadata.mastra.suspendPayload.
app/page.tsx
useInterrupt({
  render: ({ interrupt, event, resolve }) => {
    const payload = interrupt
      ? interrupt.metadata?.mastra?.suspendPayload
      : JSON.parse(event.value).suspendPayload;
    return (
      <div>
        <p>Pick a time for {payload?.topic}</p>
        <button onClick={() => resolve({ chosen_time: "Tomorrow 10:00 AM" })}>
          Tomorrow 10:00 AM
        </button>
        <button onClick={() => resolve({ cancelled: true })}>Cancel</button>
      </div>
    );
  },
});

Whatever you pass to resolve(...) arrives in the tool as resumeData, so shape it to match the tool's resumeSchema.

On the frontend, register an interrupt renderer with useInterrupt. When the agent pauses, your component mounts inline in the chat, captures the user's choice, and resumes the run with that input. The excerpt imports the demo's TimePickerCard, its fallback slots, and its suggestions. Use your own picker component in their place.

page.tsx
import {
  CopilotKit,
  CopilotChat,
  useInterrupt,
} from "@copilotkit/react-core/v2";
import type { TimeSlot } from "./_components/time-picker-card";
import { TimePickerCard } from "./_components/time-picker-card";
import { generateFallbackSlots } from "../_shared/interrupt-fallback-slots";
import { useGenUiInterruptSuggestions } from "./suggestions";

export default function GenUiInterruptDemo() {
  return (
    <CopilotKit runtimeUrl="/api/copilotkit" agent="gen-ui-interrupt">
      <div className="flex justify-center items-center h-screen w-full">
        <div className="h-full w-full max-w-4xl">
          <Chat />
        </div>
      </div>
    </CopilotKit>
  );
}

// Shape the backend `schedule_meeting` tool suspends with (its suspendSchema),
// wrapped by the @ag-ui/mastra bridge under `mastra_suspend`.
type SuspendPayload = {
  topic?: string;
  attendee?: string;
  slots?: TimeSlot[];
};

function Chat() {
  useGenUiInterruptSuggestions();

  // Native interrupt path (OSS-383). The backend `schedule_meeting` tool
  // `suspend()`s; the @ag-ui/mastra bridge surfaces that as an AG-UI interrupt
  // and `useInterrupt` renders the picker inline. `resolve(...)` resumes the
  // Mastra run (re-invoking the tool's `execute` with the selection as
  // `resumeData`).
  useInterrupt({
    agentId: "gen-ui-interrupt",
    renderInChat: true,
    render: ({ event, resolve }) => {
      // Mastra wraps the suspend value as
      // `{ type: "mastra_suspend", toolName, suspendPayload, ... }` and the
      // AG-UI adapter JSON-stringifies it — parse, then read `suspendPayload`
      // (NOT the raw value, which is the wrapper).
      const raw = event.value ?? {};
      const parsed = (typeof raw === "string" ? JSON.parse(raw) : raw) as {
        suspendPayload?: SuspendPayload;
      } & SuspendPayload;
      const payload: SuspendPayload = parsed.suspendPayload ?? parsed;
      const slots =
        payload.slots && payload.slots.length > 0
          ? payload.slots
          : generateFallbackSlots();
      return (
        <TimePickerCard
          topic={payload.topic ?? "a call"}
          attendee={payload.attendee}
          slots={slots}
          onSubmit={(result) => {
            // Defer resolve so React commits the picked/cancelled badge before
            // useInterrupt clears the interrupt element (a single rAF is not
            // reliable — it can fire before React's commit).
            setTimeout(() => resolve(result), 500);
          }}
        />
      );
    },
  });

On the backend, the agent pauses the run with its framework's interrupt primitive and waits for the user's answer before it continues.

index.ts
/**
 * Scheduling agent for the interrupt-adapted demos (gen-ui-interrupt,
 * interrupt-headless).
 *
 * This agent powers the NATIVE interrupt path (OSS-383). The backend
 * `schedule_meeting` tool `suspend()`s with a time-picker payload; the
 * @ag-ui/mastra v1 bridge maps that to an AG-UI interrupt (legacy
 * `on_interrupt` CUSTOM event + the standard `RUN_FINISHED` interrupt-outcome,
 * on by default). The frontend `useInterrupt` (gen-ui-interrupt, in-chat) /
 * hand-rolled headless subscription (interrupt-headless, app-surface) renders
 * the picker and resolves it, which resumes the run — re-invoking the tool's
 * `execute` with `resumeData`. Replaces the prior `useHumanInTheLoop`
 * frontend-tool workaround.
 *
 * Resume requires instance `storage` (see src/mastra/index.ts) so the
 * suspended agentic-loop snapshot can be reloaded.
 */
export const interruptAgent = new Agent({
  id: "interrupt-agent",
  name: "Interrupt Agent",
  tools: { schedule_meeting: scheduleMeetingInterruptTool },
  model: openai("gpt-6-luna"),
  instructions: `You are a scheduling assistant. Whenever the user asks you to book a call or schedule a meeting, you MUST call the \`schedule_meeting\` tool. Pass a short \`topic\` describing the purpose of the meeting and, if known, an \`attendee\` describing who the meeting is with.

The \`schedule_meeting\` tool surfaces an interactive time-picker to the user and pauses until they pick a slot (or cancel), then returns their selection to you. After it returns, briefly confirm whether the meeting was scheduled and at what time, or note that the user cancelled. Do NOT ask for approval yourself — always call the tool and let the picker handle the decision.

Keep responses short and friendly. After you finish executing tools, always send a brief final assistant message summarizing what happened so the message persists.`,
  memory: new Memory({
    storage: new LibSQLStore({
      id: "interrupt-agent-memory",
      url: WORKING_MEMORY_DB_URL,
    }),
    options: {
      workingMemory: {
        enabled: true,
        schema: AgentState,
      },
    },
  }),
});

Ask the agent to book a call, for example Book an intro call with the sales team to discuss pricing. The time picker appears in the chat. Pick a slot, and the agent confirms the booked time.

How is this guide?