Hanami

Private screening · public proof

A whitelist gate that can show its work.

Hanami turns your project criteria into an ERC-7857 AI bouncer, certifies it against eight fixed safety scenarios, and interviews applicants inside a TEE on 0G.

Safety gate
Strict 8 / 8
Inference
TEE verified
Artifacts
0G Storage
Ownership
ERC-7857
8 / 8
certified

Portrait / seal · activate the card to inspect

The proof ledger

Trust is a chain of evidence, not a badge.

Each layer has one job. Together they make private screening reproducible without publishing private criteria or simulated conversations.

010G Compute

TEE-attested screening

Every bouncer response must arrive with verified enclave evidence.

02Safety report

Eight fixed simulations

Thoughtful, low-effort, jailbreak, and edge-case applicants must all resolve correctly.

030G Storage

Content-addressed evidence

Private intelligence and public-safe report artifacts are pinned by root.

040G Chain

Verifiable decisions

The ERC-7857 bouncer and final applicant verdicts have durable on-chain receipts.

Owner workflow

From private intelligence to a portable whitelist.

  1. 01

    Define the gate

    Write the voice, values, and private criteria that make this community specific.

  2. 02

    Certify the bouncer

    Sign one gasless authorization. Hanami runs eight TEE-attested simulations against the exact current text.

  3. 03

    Mint in three clear steps

    Mint the iNFT, authorize the bouncer, then create the campaign. Completed transactions are never repeated.

  4. 04

    Interview and export

    Applicants complete a private 3–6 turn interview. Owners review outcomes and export a Merkle root.

Start with your campaign