The fix everyone reaches for doesn't error. It just doesn't book anything.

ON CONFLICT DO NOTHING against an exclusion constraint succeeds, inserts nothing, and raises no error at all -- a handler that doesn't check rowCount tells the customer 'booked' while the database holds no booking.

This is one of the things Scheduling & Availability Core already handles. Bookings and availability that survive timezones, repeating events and daylight saving — without double-booking.

Buy for $993 of 3 clean-room builds passed · full refund if it fails on your machine

Is this you?

The reflex fix for a 23P01 exclusion-violation error -- add ON CONFLICT DO NOTHING -- does eliminate the error, and it does so by silently inserting nothing at all, with no exception raised. A handler that doesn't explicitly check the statement's rowCount reports success to the customer while the database genuinely holds no booking row for them, which is a worse failure than the original error: it's invisible in every log line.

Why this one is easy to get wrong

ON CONFLICT DO NOTHING is the standard, idiomatic Postgres idempotency pattern for a unique-constraint violation, where it behaves exactly as expected -- and reaching for it the moment a booking insert throws 23P01 is a reasonable generalization from that pattern. What's specific to exclusion constraints, verified directly on PostgreSQL 18.4, is that the no-target form succeeds silently, while ON CONFLICT (cols) DO NOTHING instead raises 42P10 (column inference can't match an exclusion constraint) and DO UPDATE raises 42809 -- there is no upsert path for an exclusion constraint at all, in either direction.

What you get instead

The shipped booking path never uses ON CONFLICT against the exclusion constraint at all -- book() lets the plain INSERT run, catches 23P01 explicitly, and maps it to a graceful slot_taken result; a genuine 23505 unique-constraint hit (the idempotency key, which is a usable arbiter) is handled separately by re-reading the existing booking. The harness asserts all five documented behaviors of the ON CONFLICT trap directly against the pinned PostgreSQL 18.4 image, rather than assuming any one of them.

Source: ARCHITECTURE.md 'The ON CONFLICT trap (verified on PostgreSQL 18.4, not inferred)' — 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 22 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 passed

We 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 →

Buy for $99 14-day refund if verify.sh fails →