Analysis updated 2026-07-25
Study how order batching can hide individual trade direction on a public blockchain.
Run the local proof script to test the full order-pooling flow on a local validator.
Verify published Devnet evidence using the included verification script.
Experiment with threshold-based settlement using three designated approvers.
| gabchess/kageb | abc3dz/mixxx | abyo-software/ferro-stash | |
|---|---|---|---|
| Stars | 1 | 1 | 1 |
| Language | Rust | Rust | Rust |
| Setup difficulty | hard | moderate | moderate |
| Complexity | 4/5 | 2/5 | 4/5 |
| Audience | researcher | general | ops devops |
Figures from each repo's GitHub metadata at analysis time.
Requires a Solana toolchain, local validator, and familiarity with on-chain program deployment and Devnet configuration.
KageB is a prototype built on Solana that lets traders submit orders without exposing their individual choices to the public. Four real orders go into a pool, opposing orders cancel each other out internally, and only the net aggregate result reaches the market. The public settlement trace shows the group outcome, not which participant chose to buy or sell. The name comes from a Naruto technique called Kage Bunshin. The idea is that each encrypted order is like a sealed scroll in a crowd. The blockchain sees the crowd move but cannot read the choice inside each scroll. Unlike the anime technique, KageB does not create fake or decoy trades. It only pools real orders and sends the net result onward. Each participant funds both sides of a possible trade before submitting an encrypted order. This two-sided funding prevents the public funding step from revealing whether the trader intends to buy or sell. The system requires at least four valid participants before it will lock the group. Three designated approvers called keypers must reach a two-of-three threshold to authorize the lock and later the settlement. Opposing orders cancel inside the pool, and only the residual aggregate moves to a fixed venue. The prototype runs on Solana's Devnet using synthetic, zero-decimal assets. It is custodial, upgradeable, and explicitly not safe for real funds. It has no deposit, withdrawal, or refund mechanism. Each order is limited to exactly one lot at one operator-set price. The project provides a local proof script that builds the programs, starts a local validator, and checks the full flow, plus a separate script to verify published Devnet evidence. The workspace exposes Rust protocol building blocks and a reference implementation, not a hosted service or complete trading SDK. The proof does not establish unique humans, production anonymity, private funding, protection from the operator, or safe use with real assets. If an observer already knows every other participant's order, the aggregate can reveal the remaining one. Crowd size is a demo condition, not an anonymity guarantee. The full README is longer than what was shown.
A prototype on Solana that groups four real trading orders together, cancels opposing ones internally, and only sends the net result to market so individual choices stay hidden from the public.
Mainly Rust. The stack also includes Rust, Solana, Solana Devnet.
Setup difficulty is rated hard, with roughly 1h+ to a first successful run.
Mainly researcher.
This repo across BitVibe Labs
Verify against the repo before relying on details.