TukarDocumentationSCF submission materials

20 questions the deck and demo don't answer

These are the gaps. The 9-slide deck and the demo video cover the problem, the moat, the market, the tech depth, and the on-chain verification. They do NOT cover team, traction, regulatory operations, key governance, scale, economics, or recovery. A sharp judge goes straight here. Short answers and angles below. The three team and traction ones you must fill in yourself, do not fabricate.


Team and traction (fill these in yourself)

1. Who is on the team, and why are you the ones to build this? Your answer. Name the people and the one credential each that de-risks this (ZK, Stellar, payments, or compliance). If you are early, own it and pivot to the working testnet build.

2. Do you have any users, pilots, or letters of intent yet? Your answer, honestly. If none, say "not yet, it's a testnet build," then pivot to the ask: one licensed-anchor pilot is exactly what we're raising for. Do not imply traction you don't have.

3. What have you shipped before that shows you can pull this off? Your answer. Point to the fact that this is already 8 circuits, 15 contracts, 282 webapp tests and 333 Cargo tests live on testnet, which is the strongest evidence in the room.

Regulatory operations

4. The deny-list is 8 addresses. Real sanctions lists are thousands and change daily. How do you keep it current on-chain? The 8 are a demonstration set. In production the operator syncs a maintained list (OFAC, UN) into the on-chain allow-root and deny-set on a schedule, ideally through a licensed compliance-data provider. Real-time global coverage is a data-ops problem, not a circuit problem, and it's a reason to partner with a compliance vendor rather than roll our own.

5. How do you meet the Travel Rule and AML obligations for a real corridor? The licensed anchors at the edges do the KYC and Travel Rule, that is their existing job. We give them the private settlement plus the compliance proof and selective disclosure so they can meet their obligations without the public ledger leaking every customer. We are the rail, not the KYC layer. The deepening is built. A verified disclosure maps to an IVMS101 payload and goes out over a real OpenVASP TRP 3.2.1 exchange, so two anchors exchange the required originator and beneficiary data VASP-to-VASP without leaking the public payment graph: base58 Travel Addresses, canonical JSON, detached Ed25519 signatures verified on receipt with optional peer-key pinning, a tracked request lifecycle, and national identification filled from a live GLEIF LEI lookup. A TRISA companion node ships alongside it and needs the operator to register a test VASP before that leg turns on. On the KYC side we compose rather than re-do: Reclaim (zkTLS proof-of-personhood, live on Stellar) populates the ASP allow-list, because its proof is bound to the Stellar address server-side before verification. idOS (reusable KYC, live on Stellar) is integrated and verifies a real shared credential, but it cannot populate the allow-list: idOS keys a credential by its owner's idOS user id, and the consumer SDK exposes no user-keyed wallets read, so the share cannot be tied to a Stellar address.

6. If a court orders you to reveal a specific user's transactions, can you? Two layers. The licensed anchor knows the KYC'd identity at the edge. And the regulator can register an audit request the holder is required to answer, which the contract enforces. Honestly, ultimate enforcement rests on the anchor and the legal process, the protocol makes disclosure possible and provable, it does not replace a subpoena.

7. Who curates the allow-list, and what stops it becoming a censorship tool? A licensed compliance operator curates it, and yes, that is a deliberate gatekeeping point, it is exactly what separates us from a permissionless mixer. Rather than build KYC ourselves, we compose a reusable rail already live on Stellar: Reclaim zkTLS proof-of-personhood is wired and does populate the allow-list, so membership traces back to an existing verified credential rather than a list we invent. The proof is bound to the Stellar address server-side before it is verified, and a successful verification computes the new allow-root plus the exact set_asp_root invocation the operator signs. idOS reusable KYC is integrated alongside it but cannot feed the allow-list, because a consumer cannot map an idOS user id to that user's registered wallets, so a verified share cannot be tied to a Stellar address. Governance of the list is a real production question, and it is the kind of thing an anchor partner and a regulator would sign off on, not us alone.

Security and keys

8. Who holds the pool's admin keys? What if the operator is malicious? There is an operator key that sets policy, and admin writes are signed offline, never in the browser. In the demo it's one key, in production it should be multisig or governance. A malicious operator can change policy but cannot forge a proof or take escrowed funds outside the contract's rules. Key governance is a named hardening step.

9. What is your trusted-setup story, and what if it was compromised? Groth16 needs a per-circuit setup. Phase one is the Hermez powers-of-tau the whole ecosystem trusts. Phase two is our own ceremony, three contributions plus a public beacon, byte-identical to the committed transcripts. The honest gap is we ran the rounds on one machine, so genuinely independent contributors are a first funded step.

10. It's not audited. What is the contract attack surface? Correct, not professionally audited yet. It is hardened through many adversarial self-review rounds against a documented threat model, with 333 passing Cargo tests across the contract crates (52 of them on the live pool) and 230 passing webapp tests. A professional audit is the first use of prize or grant money.

11. What stops metadata leakage, linking a deposit to a withdrawal by timing or amount? The crossing hides the amount and both parties, but deposits and withdrawals are public at the edges by design, so linkability shrinks as the anonymity set grows. Timing and amount correlation at the edges is a known limitation, and mitigations like fixed denominations and settlement delays are on the roadmap.

Scale and economics

12. How many transactions per second, and what are the proving and tree limits? Proving is a few seconds client-side, which is fine for the act of sending money. The tree is depth 10 today and extends. Throughput is bounded by Soroban, not by us, and is not a bottleneck at pilot scale. Deeper trees and batching are the scale levers.

13. What is the on-chain cost per transfer, and who pays gas? Soroban fees are a fraction of a cent, the pairing verification is the main cost. The user pays, or the fee can be sponsored by a relayer through Stellar's native fee-bump, which we have already proven on testnet.

14. What are the unit economics and margins? A thin take-rate, a few basis points on settlement volume, paid by the anchors. Once an anchor is integrated the marginal cost is software, so margins are strong, the real variable is corridor volume.

15. How big does the anonymity set need to be for the privacy to be real? Meaningfully private means hundreds or more per epoch sharing a pool, and you get there by a distribution partner driving real volume into shared pools. Demo scale is small, we say so in the app and surface the set size so no one is misled.

Product, go-to-market, and recovery

16. Which exact corridor and which anchor first, and why that one? Your call to name it, the logic is a high-volume, high-cost lane where a licensed anchor already operates, so the fee pain is largest and the fiat edges already exist.

17. What stops the anchor from cutting you out and building this themselves? And how are you different from the other privacy pools (Wraith, AnchorShield, Moonlight)? The moat is the compliance-and-privacy stack, eight circuits, four disclosure types, the audit registry, which is a year of work, not an anchor product team's afternoon. And we position Tukar as the neutral compliant settlement layer that anchors plug into, one policy per corridor, across many anchors, so no single one owns it and rival privacy apps become integrators rather than competitors. On the crowded field: Wraith and AnchorShield won at this hackathon and Moonlight is a live SCF-funded confidential-transactions project, all on the compliant-privacy-pool idea, so we differentiate on being a remittance corridor with real fiat edges, an oracle-gated off-ramp, and the anchor-layer packaging, plus a built compliance layer a generic pool does not touch: a real OpenVASP TRP 3.2.1 Travel Rule exchange with verified signatures, Reclaim proof-of-personhood feeding the allow-list, cryptographic proof-of-reserves, and an on-chain per-corridor policy registry. It is a real risk we answer by moving fast and staying multi-anchor.

18. Is there a token? No. Revenue is fees on settlement. No token keeps it simple and avoids a whole layer of regulatory complexity.

19. What is the 6-month plan? Audit, then a genuinely distributed trusted-setup ceremony, then one licensed-anchor pilot live on one corridor, and measure real usage and privacy-set growth. That is also the Stellar Community Fund milestone story.

20. What happens if a user loses their claim note or a transaction fails? The claim note is a bearer instrument, so losing it is like losing cash, which is a real UX risk, and production adds an optional claim-to-address mode or custodial recovery. A failed transaction doesn't lose funds, the deposit is atomic on-chain and the escrow releases only under the contract's rules.


The move when you get one you can't fully answer

Name the gap honestly, say what you'd do about it, and tie it to the ask. "That's exactly the kind of thing the pilot and the audit money is for." Judges trust a team that knows its own gaps more than one that pretends there are none.