Dr Roosh

I went looking for how Citadel actually handles a revoked credential. Turns out the answer isn't "the chain checks it." Dusk Network's Citadel protocol $DUSK #dusk lets a user prove a session is cryptographically valid — that a real License Provider signed a real license — but per the docs, that proof does not decide service policy.
That call sits with the Service Provider. Per @Duskfoundation's own documentation, the SP decides which License Providers it trusts, which attributes it accepts, whether a session is expired or revoked, and whether the session cookie can be reused. None of that is written into the on-chain verification itself.
What changed for me was realizing "privacy-preserving KYC" here doesn't mean the chain enforces compliance. Personal attributes are never written to the blockchain — that part's explicit in the docs. But expiry, revocation, and trust in the issuer are policy decisions each Service Provider makes independently, off-chain, with nothing on-chain forcing consistency between them.
@Dusk

Attendez — le “mécanisme” qui protège actuellement votre transaction DuskEVM contre les bots n’est même pas cette technologie sophistiquée.
Tout le monde parle de Hedger, le moteur de confidentialité de DuskEVM ($DUSK #dusk @DuskFoundation) — chiffrement homomorphe et preuves ZK, soldes chiffrés, vraie cryptographie. Pourtant, ce n’est pas ça qui empêche le front-running aujourd’hui.
J’ai vérifié la documentation. DuskEVM exécute actuellement uniquement un séquenceur. Pas de mempool publique. Un seul séquenceur, donc rien à surveiller pour que des bots puissent “passer devant”. C’est le mécanisme réel.
Il y a donc deux récits de “confidentialité” différents superposés, et il est facile d’attribuer le bon modèle au mauvais. Hedger, lui, correspond à une vraie confidentialité cryptographique, pensée pour être conforme aux audits. La protection contre le front-running, c’est autre chose — un effet secondaire du fait d’avoir un seul séquenceur sans rien d’exposé.
Voici ce que je n’ai pas encore trouvé : une documentation ou une feuille de route indiquant si le séquenceur de DuskEVM reste unique ou s’ouvre avec le temps. S’il devient un système multi-parties ou public, cette protection précise devra être remplacée. Rien n’est confirmé à ce sujet — c’est simplement une question que soulève l’installation actuelle.
Je suis curieux : quelqu’un a-t-il déjà suivi la feuille de route réelle du séquenceur de DuskEVM ? Honnêtement, je n’ai pas la réponse.
@Dusk

Spent the afternoon poking around Dusk's repos instead of just reading the pitch deck. #Dusk $DUSK @DuskFoundation — "privacy on a public chain" is the whole TradFi hook, so I wanted to see what's actually moving, not what's being said.
Here's the thing that stuck: duskevm-genesis, the repo holding the genesis block and rollup config, had commits land Aug 8 and Aug 10, 2026. Not RWA settlement code. Not a securities contract. Genesis and rollup plumbing — the boring, load-bearing stuff nobody screenshots.
Meanwhile DUSK's most active pair, DUSK/USDT on Binance, was doing about $117k in 24h volume when I checked, against roughly $3.07M total across 51 markets (CoinGecko). For a project whose entire thesis is "institutions will finally transact on a public chain," that's a thin signal.
Made me reconsider my own framing going in — I assumed compliant privacy meant invisible institutional flow humming in the background. What I found instead was infra still being wired, quietly, while the market cap sits under $50M.
So which comes first here — the institutions Dusk keeps promising, or the plumbing finally catching up to the pitch?
@Dusk

Reading the Phoenix circuit constraints stopped me. On Dusk $DUSK #dusk @Duskfoundation the Transfer Contract never looks at the actual note values or which Merkle leaf is spent.
It only checks a PLONK proof that the private inputs satisfy five conditions: the note hash opens against a recent Merkle root of notes, the prover knows the note secret key, the nullifier equals Poseidon(npk′ ‖ position), the output commitments open correctly, and the sum of input values equals outputs plus fee plus any deposit.
Nullifiers themselves are published so the network can reject reuse. Because each is derived from the hidden note key and position, no observer can map a nullifier back to a specific leaf. Validity lives entirely inside circuit satisfaction; the contents never appear on the ledger.
What changed for me was realizing double-spend protection and balance integrity both sit inside the proof rather than any visible state change.
Next check: whether every accepted Phoenix transaction’s published nullifiers stay unique in the on-chain nullifier set after finality.
@Dusk