All packsWebhook Durability & ReplaySolve
webhook worker crash retry attempts counter order
Increment the retry counter at claim time, or a crash loses the count
Incrementing a delivery's attempts counter after processing, instead of at the moment a worker claims the row, loses the count whenever a worker is killed mid-delivery (D11).
This is one of the things Webhook Durability & Replay already handles. Receive webhooks without losing, duplicating or misordering them when the sender retries.
Is this you?
Incrementing attempts only after a delivery finishes processing means a worker killed mid-delivery -- a deploy, an OOM, a crash -- loses that attempt entirely: the row goes back to pending with its counter unchanged, and a delivery that should be approaching the dead-letter threshold quietly gets more retries than it was ever supposed to have.
Why this one is easy to get wrong
Counting an attempt once the work is actually done reads as the more accurate way to track retries -- why count something that might still succeed? The gap only appears at the exact moment that matters most: a worker crash mid-processing, which by definition happens without any code afterward getting a chance to record that the attempt occurred.
What you get instead
D11 increments attempts inside the same UPDATE that claims the row for processing (D9), before any handler logic runs at all -- so a claimed row already carries the attempt it's about to make, and a row dies at exactly attempts = WEBHOOK_MAX_ATTEMPTS regardless of how many times a worker was killed mid-delivery. backoffMs(attempt) is a pure function over a fixed table with deterministic-seeded jitter, existing only to stop a restarted worker from re-hitting a struggling dependency in lockstep with every other restarted worker.
Source: ARCHITECTURE.md D11 — checkable in the pack you receive
How you actually use this
You don’t install a library or wire up an SDK. Your own coding agent builds the code in your project, and you keep it — no runtime dependency on us.
Step 1
Download and unzip
You get a folder: the docs that tell an agent what to build, a starting skeleton, and the test suite that decides when it's done.
Step 2
Open it in Claude Code or Cursor
Point your coding agent at the folder. Nothing to install, no account with us, no API key.
Step 3
Paste one prompt
The pack contains the exact prompt. Paste it as your first message and leave it alone — it works through the build itself, choosing a cheaper or stronger model per task.
Step 4
Run ./verify.sh
One command. It prints a pass or fail for every check. Green means the build is done — the same script we ran to produce the receipt on this page.
Typical build: about 40 minutes of your agent working, mostly unattended. Then you integrate the working module into your app the way you would any code you’d written yourself.
Why you can believe this
3 of 3 runs passedWe ran this pack from an empty folder 3 times and published exactly what happened — every check, the model, the token cost, the wall time. Not a testimonial, and not our opinion: the same verify.sh you run yourself. Read the full receipt →