smee.io alternatives

smee.io is the webhook-proxy the Probot/GitHub Apps world standardised on: create a channel, point GitHub at it, run the Node smee-client, and events stream to your localhost. It's free and it works — until you hit its structural limits. If you're searching for an alternative, it's probably one of these:

Dropped deliveries. smee pushes events over a live connection. If your client isn't running — laptop asleep, connection silently dropped, you just hadn't started it yet — those webhooks are gone. There's no durable queue to catch up from.
No real inspection or history. You get the payload forwarded, but there's no persistent capture log to search, diff, replay, or export later.
Channels are unauthenticated. Anyone who has (or guesses) your channel URL can subscribe to your payloads. Fine for toy repos; uncomfortable for anything real.
It needs Node. The official client is an npm package — annoying in a Go/Python/Rust project or a slim container.

1. Hookden relay (this site)

$ curl -s https://hookden.pages.dev/cli -o hookden && chmod +x hookden
$ ./hookden new                              # prints your webhook URL
$ ./hookden relay YOUR_BIN http://localhost:3000

Nothing is ever dropped. The capture URL is a durable bin: webhooks land and persist whether or not the relay is running. Start the relay after lunch and pick up where you left off — or relay --all to re-deliver the whole history.
Full inspection for free: every delivery is also in the web dashboard — headers, JSON pretty-printed, ✓/✗ HMAC signature badges (GitHub/Stripe/Shopify/generic), diff two deliveries, copy as curl, replay, HAR/JSON export.
No Node, no npm. The client is a POSIX shell script that needs only curl and sed. Works in any container that has curl.
Signatures still verify: bodies are re-delivered byte-exact, so X-Hub-Signature-256 checks pass in your local handler. Full walkthrough.

Trade-offs, honestly: relay polls, so deliveries reach localhost within a couple of seconds rather than push-instantly; bodies over 100 KB are truncated on relay (the CLI warns you, since a truncated body breaks signature checks); and it's hosted — if payloads can't leave your network, self-host instead.

2. Self-host smee

Good: the smee server is open source; running your own instance fixes the public-channel concern.
Trade-offs: you still have the no-persistence, drop-on-disconnect model, the Node client, and now a service to operate.

3. A tunnel (ngrok, cloudflared, localtunnel)

Good: real-time, works for any HTTP traffic, not just webhooks.
Trade-offs: a daemon exposing a port on your machine, URLs that rotate on restart (free tiers), auth walls and limits. If the traffic is only webhooks, a tunnel is more machinery than the job needs — honest tunnel comparison here.

Quick chooser

Webhooks to localhost, nothing lost, inspectable history → Hookden relay.
You're all-in on Probot and never disconnect → smee is fine, keep it.
Payloads must stay in-house → self-host smee or a tunnel to a private box.

No signup needed. Or from your terminal: curl https://hookden.pages.dev/new

← All guides · Docs · Hookden vs webhook.site