Webhooks & External Callbacks

Receive HTTP callbacks from external services, process them durably, and respond inline.

Use webhooks when external services push events to your application via HTTP callbacks. The workflow creates a webhook URL, suspends with zero compute cost, and resumes when a request arrives.

Accepting callbacks from payment processors (Stripe, PayPal)
Waiting for third-party verification or processing results
Any integration where an external system calls you back asynchronously

Create a webhook with manual response control, then iterate over incoming requests:

import { createWebhook, type RequestWithResponse } from "workflow";

declare function processEvent(request: RequestWithResponse): Promise<{ type: string }>; // @setup

export async function paymentWebhook(orderId: string) {
  "use workflow";

  const webhook = createWebhook({ respondWith: "manual" }); 
  // webhook.url is the URL to give to the external service

  const ledger: { type: string }[] = [];

  for await (const request of webhook) { 
    const entry = await processEvent(request);
    ledger.push(entry);

    // Break when we've received a terminal event
    if (entry.type === "payment.succeeded" || entry.type === "refund.created") {
      break;
    }
  }

  return { orderId, webhookUrl: webhook.url, ledger, status: "settled" };
}

Step function for processing

Each webhook request is processed in its own step, giving you full Node.js access for validation, database writes, and responding to the caller:

import { type RequestWithResponse } from "workflow";

async function processEvent(
  request: RequestWithResponse
): Promise<{ type: string }> {
  "use step";

  const body = await request.json().catch(() => ({}));
  const type = body?.type ?? "unknown";

  // Validate, process, and respond inline
  if (type === "payment.succeeded") {
    // Record the payment in your database
    await request.respondWith(Response.json({ ack: true, action: "captured" })); 
  } else if (type === "payment.failed") {
    await request.respondWith(Response.json({ ack: true, action: "flagged" }));
  } else {
    await request.respondWith(Response.json({ ack: true, action: "ignored" }));
  }

  return { type };
}

Submit a request to an external service, pass it your webhook URL, then race the callback against a deadline:

import { createWebhook, sleep, FatalError, type RequestWithResponse } from "workflow";

export async function asyncVerification(documentId: string) {
  "use workflow";

  const webhook = createWebhook({ respondWith: "manual" });

  // Submit to vendor, passing our webhook URL for the callback
  await submitToVendor(documentId, webhook.url);

  // Race: wait for callback OR timeout after 30 seconds
  const result = await Promise.race([ 
    (async () => {
      for await (const request of webhook) {
        const body = await processCallback(request);
        return body;
      }
      throw new FatalError("Webhook closed without callback");
    })(),
    sleep("30s").then(() => ({ status: "timed_out" as const })), 
  ]);

  return { documentId, ...result };
}

async function submitToVendor(documentId: string, callbackUrl: string): Promise<void> {
  "use step";
  await fetch("https://vendor.example.com/verify", {
    method: "POST",
    body: JSON.stringify({ documentId, callbackUrl }),
  });
}

async function processCallback(
  request: RequestWithResponse
): Promise<{ status: string; details: string }> {
  "use step";
  const body = await request.json();
  await request.respondWith(Response.json({ ack: true }));
  return {
    status: body.approved ? "verified" : "rejected",
    details: body.details ?? body.reason ?? "",
  };
}

When payloads are too large to serialize into the event log, pass a lightweight reference (a "claim check") instead. Use a hook to signal when the data is ready:

import { defineHook } from "workflow";

export const blobReady = defineHook<{ blobToken: string }>(); 

export async function importLargeFile(importId: string) {
  "use workflow";

  // Suspend until the external system signals the blob is uploaded
  const { blobToken } = await blobReady.create({ token: `upload:${importId}` }); 

  // Process by reference -- the full payload never enters the event log
  await processBlob(blobToken);

  return { importId, blobToken, status: "indexed" };
}

async function processBlob(blobToken: string): Promise<void> {
  "use step";
  // Fetch the blob using the token, process it
  const res = await fetch(`https://storage.example.com/blobs/${blobToken}`);
  const data = await res.arrayBuffer();
  // Index, transform, or store the data
}

Resume from an API route when the upload completes:

import { resumeHook } from "workflow/api";

// POST /api/upload-complete
export async function POST(request: Request) {
  const { importId, blobToken } = await request.json();
  await resumeHook(`upload:${importId}`, { blobToken }); 
  return Response.json({ ok: true });
}

respondWith: "manual" gives you control over the HTTP response from inside a step. Use this when you need to validate the request before responding.
for await on a webhook lets you process multiple events from the same URL. Use break to stop listening after a terminal event.
Webhooks auto-generate URLs at /.well-known/workflow/v1/webhook/:token. Pass this URL to external services.
Race webhooks against sleep() for deadlines. If the callback doesn't arrive in time, the workflow can take a fallback action.
For large payloads, use a hook + reference token instead of passing the data through the workflow. The event log serializes all step inputs/outputs, so large payloads hurt performance.

"use workflow" -- marks the orchestrator function
"use step" -- marks functions with full Node.js access
createWebhook() -- creates an HTTP endpoint the workflow can await
defineHook() -- creates a typed hook for signal-based patterns
sleep() -- durable timer for deadlines
FatalError -- prevents retry on permanent failures

推荐订阅源

Workflow SDK Documentation

Step function for processing