Proof #0001 · PASS

Engine calibration on an already open program: SPL Memo v3, rebuilt source. Not a cracked closed program.

MemoSq4gqABAXKb96qnH8TysNcWxMyWCqXgDLGmfcHr

Match (compared cases)

10,985/10,985

11,000 cases, 15 budget-bound

Real mainnet instructions

1,000/1,000

plus 9,985 fuzzed compared, seed c34cf12a2e0f0762

Success path on the original

7,107

at least 100 required

Mismatches

all 10985 compared cases match, 7107 on the success path

Hashes

Report sha256

4b029ffb…0f0dab03

Source commitment

ca60fa3a…2dc838d2

Corpus bundle

731ba8d3…47a17858

Original binary

f520eaf0…6ef67954

Candidate binary

dafe4811…3f52b6f9

Outcomes, original

6979bdf2…9217476f

Outcomes, candidate

32b9ede4…103e0ead

Judged

2026-11-02 15:44 UTC

Memo record

decomp:proof:calibration:mvzfiv4jc190a923:ca60fa3a2a704639c9ee01f94f80871e4bdc4ae978b2d8116cfdaa4b2dc838d2:4b029ffbe428ca3d614c609b9bd56e5379afb1193b9ed341e56ded470f0dab03:10985/10985

Memo record pending: the keeper records it on Solana once it is funded.

Download the source

curl -fsSL "$SITE/verifier/verify.sh" | SITE="$SITE" sh -s -- 0001

Re-run this verdict →

First mismatches

Every compared case matched: result, logs line by line, return data, and every writable account.

Compared: result and error, logs line by line, return data, lamports owner and data of every writable account. Stripped: lines 'Program <id> consumed N of M compute units'; cases where the original exhausts its compute budget are counted as budget_bound and not compared. decomp-replayer 1.0.0, LiteSVM 0.17.0, litesvm mainnet feature set.