𝑹𝒂𝒆𝒆🦉

This one is about one concrete thing you can look at today: the @primus_labs Confidential Vault on BNB Chain, built with Unitas and United Stables. The problem it targets Public vaults show everything. anyone can see your deposit size and track a position. for an individual that's an annoyance. for a treasury or fund it's a cost, because a visible position can be followed, copied or traded against. The design: a second door into the same vault, the confidential vault is an encrypted deposit option added to an existing, public lending vault. Depositors get the same strategy, the same risk profile, the same collateral basket and the same yield as the public version. what changes is that deposit sizes are hidden onchain. That is a deliberately modest idea. It doesn't invent a new yield source. It changes who can see you. How it works, step by step 1) Shield Convert a supported stablecoin into its confidential form in one transaction. the confidential token wraps a standard dollar-pegged one and is built with FHE, so balances and transfer amounts stay encrypted onchain. 2) Stake Deposit the confidential stablecoin into the confidential vault. 3) Earn Yield accrues through the underlying lending market, the same as a public deposit. 4) Unshield When you want out, convert back to the standard token. Who is involved, as far as I can tell Primus built the confidential layer with Unitas and United Stables on BNB Chain. one third-party post describes the flow as converting $U (United Stables) into an encrypted $EU, which can then be moved privately or deposited into Unitas vaults. What stays hidden and what doesn't Hidden: balances, transfer amounts and deposit sizes. per that same third-party post, individual amounts are hidden while the total batch can still be checked. that last part matters: confidentiality here is not the same as nobody being able to audit anything. How it connects to the rest of Primus Primus's pitch is confidentiality over anonymity, with selective disclosure. the vault is the FHE side of that. The zkTLS side shows up in Proof of Reserves, which is live for Unitas USDu and Perena USD* with Brevis: reserves can be verified without publishing every account. together: hide the positions, prove the backing. Honest limits • FHE carries real computation overhead, so scalability is a fair question • Confidentiality still has to coexist with regulatory auditability, and the frameworks are still developing • The model is new and hasn't been through full market cycles or stress tests • Some of the FHE stack is planned, not live, for example the threshold MPC layer • Over $1M encrypted and staked in the first 24 hours. that figure is self-reported and not audited, so treat it as a signal, not a verdict How I'd use it Read the docs first and check the contract addresses and supported assets from official sources before moving anything. Treat it as a live way to see FHE work, not as a finished product. What I find interesting: the best privacy feature may be the one that doesn't ask you to change your strategy at all. Would you use a confidential deposit for a treasury position or is public by default fine for you?

Screenshots are the default evidence on CT. PnL screenshots, dashboard screenshots, "look at my stats" screenshots. everyone knows they can be edited and everyone still treats them as proof. @primus_labs and Kaito Studio now have an integration aimed at exactly that habit, using zkTLS so users can verify off-chain activity without exposing account details or credentials. Here is what I can say with confidence, and what I can't. What's described • Users prove activity from outside the platform instead of pasting a screenshot • The proof is made without sharing a login or the full underlying data • Primus mentions Kaito's Pulse release and native browser-side verification, so users can bring authenticated platform footprints, stats and achievements into it Why this is a practical use case Reputation online is currently a claim plus a picture. a proof changes the shape of it: the claim comes with evidence that it really came from the source and the person never hands over their account. The honest limit: zkTLS proves that the data came from the claimed source. It doesn't prove the source is the whole picture. a verified stat is a verified stat, not a verdict on quality. the value is that the evidence stops being editable, not that it stops needing judgment. My view: the interesting part isn't the cryptography. It's whether a verified signal changes behavior. If creators and readers start asking "is this verified?" before "how big is it?", reputation gets harder to fake. that's the real test and it will be decided by how many people actually use it. Would you trust a verified stat over a screenshot? and what would you want verified first?

Goodnight CT. I've written a lot of explainers on zkTLS and FHE. this one is different: it's the questions I'd actually want answered if I were going to rely on this stack. Start with a concrete case. a lender wants to know whether a borrower holds more than X on an exchange. with zkTLS the borrower proves "balance above X" and the lender never sees the statement. Clean. Now the questions that come up as soon as you try to use it: 1) How fresh does a proof need to be? A proof is a moment in time. balances move. who decides how often it must be refreshed, the app, the lender or the protocol? 2) What is the proof actually worth? zkTLS proves where the data came from, not that the source is right. a proof of what an exchange displayed is only as good as that exchange. so the question is which sources an application should trust at all. 3) Who verifies the verifier and when? @primus_labs Network attestors run in TEEs, stake and face penalties, which is a sound design. penalties and governance aren't active yet and operators are whitelisted. I'd like to see the path to open participation and a rough timeline. 4) MPC or Proxy and at what cost? MPC mode is stronger against tampering. Proxy mode is faster but adds a network assumption. real numbers would help builders choose. 5) Is the friction acceptable outside crypto? Tasks today involve a browser extension, a wallet connection and sometimes a little gas on Base. that's fine for crypto-natives. what does it take for everyone else? 6) Does a verified account mean a unique person? a proof shows control of an account. It doesn't show that one person holds only one. with points and reputation in play. What I like is that none of these are reasons to dismiss the idea. they are the work between a good design and something people depend on. Teams that can say what is unfinished are the ones worth following. My view: the cryptography is the easier half. the harder half is source trust, freshness and incentives. Those are my six. Which would you add, and which would you drop? 👇

"Privacy is a human right." Tyler winklevoss said it at TOKEN2049 and it is worth taking seriously, because the hard part is not agreeing with it. the hard part is building systems that make it true. here is the tension. finance, credit and identity all need trust. trust needs evidence and evidence has usually meant handing over the whole record: bank statements, account logins, full histories. privacy loses every time that is the only option. The way out is not anonymity and it is not full transparency. It is selective disclosure: prove the fact that matters, keep everything else. Two principles @primus_labs has pointed to on this: confidentiality over anonymity and selective disclosure. What that looks like in practice 1) Prove the fact, not the record With zkTLS, a user can prove something from a web account, such as a balance threshold or an account's status, without sharing the login or the raw data. the verifier checks the proof, not the file. 2) Keep data encrypted while it is used FHE lets computation happen on encrypted data. The Confidential Vault on BNB Chain lets users shield a supported stablecoin, move or stake it and unshield it. FHETransform, which aims to turn standard EVM contracts into confidential ones, is the direction for what comes next. 3) Keep verification honest Primus Network attestors run inside TEEs, stake tokens and face penalties for wrong execution, so the checking layer has something at stake too. Honest limits • zkTLS proves provenance, not that a source is truthful • FHE carries real computation overhead and some parts of the stack are planned, not live • Attestor penalties and governance are not yet active The point stands. a right to privacy means little if the only way to participate is to give it up. Systems that let people prove what is needed and keep the rest are how the principle becomes practice. Privacy is a right. Proof is how you keep it.

🇸🇬 TOKEN2049: Gemini co-founder Tyler Winklevoss says, “Privacy is a human right.”