Skip to content

feat: first-class helper for Standard Webhooks (Svix) signature verification #4928

Description

@bobclarkdramisinfo

Title : feat: first-class helper for Standard Webhooks (Svix) signature verification

Labels : enhancement


Is your feature request related to a problem? Please describe.

I'm building event-driven automations where Trigger.dev jobs react to webhook events emitted by Composio (and a growing number of other providers that have adopted the Standard Webhooks spec — Stripe, GitHub, Slack, Linear, etc.). Composio sends three headers (webhook-id, webhook-timestamp, webhook-signature: v1,<base64>) and the signature is HMAC-SHA256 over ${id}.${timestamp}.${raw_body} with the secret base64-decoded and a tolerance window (default 300s).

The built-in verifyRequestSignature helper only covers single-header schemes (Stripe/GitHub/Cal.com/Novu style), so it can't validate Standard Webhooks signatures out of the box. I get why — supporting every provider's signature scheme isn't Trigger.dev's job, and the helper is the right escape hatch for ad-hoc cases.

The friction isn't the helper itself; it's that every Trigger.dev user integrating with a Standard Webhooks provider ends up writing the same ~50 lines of crypto + header parsing in their verify lambda. That's the part that feels like it could live in the SDK.

Describe the solution you'd like to see

A first-class helper that handles the Standard Webhooks signature scheme, roughly:

import { verifyStandardWebhooks, STANDARD_WEBHOOKS_TOLERANCE_SECONDS } from "@trigger.dev/sdk";

const endpoint = client.defineHttpEndpoint({
  id: "composio",
  source: "composio.dev",
  verify: async (request) => {
    return await verifyStandardWebhooks({
      request,
      secret: process.env.COMPOSIO_WEBHOOK_SECRET!,
      // optional, defaults to 300s per the Standard Webhooks spec
      tolerance: STANDARD_WEBHOOKS_TOLERANCE_SECONDS,
    });
  },
});

What the helper should cover:

  • Read webhook-id, webhook-timestamp, webhook-signature headers
  • Validate the v1, signature prefix (and tolerate future v1a,/v1b, variants if the spec adds them)
  • Base64-decode the secret before HMAC
  • Compute HMAC-SHA256(secret, ${id}.${timestamp}.${raw_body}) with crypto.timingSafeEqual
  • Enforce the timestamp tolerance window (anti-replay)
  • Return the same { success: true } / { success: false, reason: "..." } shape as verifyRequestSignature so it's a drop-in

Bonus if it also exposes the parsed, verified payload (mirroring what the Composio SDK does on its verifyWebhook return type) — that lets users skip the manual JSON.parse after success: true.

Describe alternate solutions

I considered a few:

  1. PR my own implementation into the SDK. I can do this, but the vouched-contributor + changesets + single-feature-per-PR process makes it heavier than a feature request first. And honestly I'd rather the maintainers pick the API shape — verifyStandardWebhooks vs. a generic verifyWebhook({ scheme: "standard-webhooks", ... }) flag, where in the helper tree it lives, etc.
  2. Publish a separate community package (@trigger-dev-community/standard-webhooks-verify). Viable, but the moment Trigger.dev ships its own helper, the community package dies. Putting it upstream also makes it discoverable next to verifyRequestSignature.
  3. Move the verification outside Trigger.dev. A small service in front of Trigger.dev that does the verification + republishes to a queue that Trigger.dev jobs subscribe to. Works, but it's exactly the kind of workaround that disappears when the gap closes. I'd rather not maintain it if I don't have to — and it pushes trust boundaries out of the SDK, which is the opposite of where they belong.

Additional information

Adoption is broader than it looks. Standard Webhooks isn't a niche format — it's becoming the de facto signature spec for SaaS webhooks:

  • Big-name migrations to Standard Webhooks: Stripe, GitHub, Slack, Linear, Shopify (via Svix)
  • Svix-native providers: most new B2B SaaS that use Svix as their delivery layer (Resend, Novu, Knock, Courier, ngrok, Composio, etc.)
  • Spec is actively maintained: the Standard Webhooks org publishes reference SDKs in Node, Python, Go, Ruby, Java, .NET, Rust, PHP — so wrapping one of those in a Trigger.dev helper is low-risk and well-tested upstream crypto

That means a single helper in the SDK would unblock integrations with every provider in the list above — not just one. The same ~50 lines of crypto currently written by hand per integration become write-once-for-everyone.

Why now : the Standard Webhooks spec keeps winning — Stripe and GitHub both migrated to it; Linear, Svix-hosted providers, and a long tail of "send us webhooks" SaaS follow. Every new Standard Webhooks integration in Trigger.dev costs ~50 lines of boilerplate that the SDK could own.

I'm not blocked on this — my workaround works — but it felt worth flagging since the cost of closing the gap is low and the gain compounds with every new Standard Webhooks integration anyone tries to build.

Spec references :

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions