One compiled trigger.
One hundred eleven real settlements.

A single compiled Stripe® trigger fanned out into real, irreversible mainnet settlements across three VM families — Base, Solana, and Hedera. Gas was abstracted away; users never touched a native token. This page shows the whole run — the fifty-three that landed on the first live fire, the bugs that stopped the rest, the fixes, and the re-fire to 111 / 111.

Nothing is hidden and nothing is projected. Every settlement below links to its public block explorer.

Trigger pi_3Tezfy…vj4U2f Mode LIVE Stripe gate PASSED · livemode=true
1
Compiled trigger
111
Settlements, on-chain
3
VM families
0
Open failures
The whole story, in five steps.

A first live run does what a first live run should: it tells you the truth. Here is the complete path from the first trigger to the final, reconciled result — every stage with its real artifacts.

Step 01 · Jun 5
First live fire
One compiled trigger, fanned out across three chains in one motion.
Step 02 · Jun 5
Fifty-three settled
Base and Hedera receipt-mints and cross-domain calls landed and confirmed.
Step 03 · Diagnosed
Four root causes
The remaining fifty-eight stopped on four small, well-understood faults — each named.
Step 04 · Fixed
Fixes shipped
All four resolved. The ledger is idempotent — it retries only what didn't land.
Step 05 · Jun 7
Re-fire to 111/111
Every remaining operation settled. The full set is reconciled and complete.
The first fifty-three.

The trigger compiled cleanly and fanned out. On the first attempt, fifty-three operations settled and confirmed on-chain — receipt-mints and cross-domain calls on Base and Hedera. These are the genuine first-run artifacts, exactly as they landed. Every hash is live on its public explorer.

Verified on the first fire

53 / 53 confirmed on-chain
Fifty-eight operations, four root causes.

The bugs surfaced only because the run was real. Each was diagnosed, named, and small. The headline finding: a single transfer-path templating fault accounted for every transfer on every chain — one fix recovered the entire category.

The third chain attempted was Solana. None of its 22 operations produced a verified settlement on the first fire — its mints and cross-domain calls hit Metaplex's 32-character name cap, and its transfers hit the same <asset> placeholder fault as every other chain. That is why the first-fire settlements show Base and Hedera only: Solana's full attempt is accounted for here in the failures, not omitted. All of it recovers on the re-fire below.

Then the other fifty-eight.

With the four fixes shipped, the trigger was re-fired. The ledger is idempotent: it skipped the fifty-three that had already landed and retried only the rest — no re-charge, no double-spend. Every remaining operation settled. Below are the fifty-eight that the re-fire recovered, grouped by chain.

Recovered on the re-fire

58 / 58 confirmed on-chain
Complete reconciled ledger — all 111 settlementsexpand to audit every hash ▾
111 / 111, reconciled.

One compiled trigger produced one hundred eleven real, irreversible settlements across three distinct virtual-machine families — a balanced fan-out of receipt-mints, native transfers, and cross-domain calls. This is the complete composition, straight from the ledger.

By chain · 111 settlements

Base · EVM
73
Solana · SVM
22
Hedera · HTS
16

By operation · 111 settlements

Receipt mints
37
Native transfers
37
Cross-domain
37

Why we published the whole run.

This was a live fire, not a rehearsal — so we are showing exactly what happened: the fifty-three that settled on the first attempt, the fifty-eight that didn't, the precise reasons why, and the re-fire that closed every one of them. Results are posted in the open for public scrutiny because infrastructure earns trust by being auditable, not by being polished.

The compilation core is deterministic, not trained — the same intent compiles the same way, every time, on every rail. The bugs that surfaced were in the surrounding execution layer, not the compiler, and they surfaced only because the run was real. We caught them on day one, named them, fixed them, and re-fired to a complete result.