QQ-VITRNPost-quantum seals

Person 1 · Midnight Preview · post-quantum signature stays in the browser · chain holds the anchor · not mainnet · Open the steps

Three people. Midnight first. Not mainnet. Writing on this page follows the person you pick.

Anchor on the selected chain

Person 1 seals job with a dual-family epoch in this browser. This button asks Midnight Preview to list one native transfer of that digest. A link appears only after the indexer lists it. Nothing here is mainnet.

Midnight Preview · job · epoch seal

Seal the job: prompt, model, output hash.

Those three lines are the statement. Both families sign them. The output bytes stay in this browser unless you put their hash in the text.

LayerWhat one break is not enough forSize
ML-DSA-65Forge if lattices hold3,309 byte signature
SLH-DSA-SHA2-128sForge if hashes hold, even if lattices fall7,856 byte signature
ML-KEM-768Read this epoch's wrap without its secret1,088 byte ciphertext
WOTS-SHA256Bind the public objects; SHA-256 steps only67 chains, one-time

A forger has to break two different assumptions. A later leak of one epoch's wrap does not open the next epoch. This is not a new hardness assumption, and it is not a claim that an AGI or a quantum computer cannot attack the code. The cut itself is not a Midnight transaction. A published sample is midnight-epoch.json.