Tuyo: Daxos Teardown

A self-custodial crypto wallet with four things bundled on top: virtual USD and EUR account numbers in the user's name, a Visa card, DeFi yield, and a discretionary card-forgiveness mechanic called Buy Now, Pay Maybe.

Tuyo builds none of the four. Fiat rails and the account numbers are Bridge. The card is issued through Signify Holdings, Inc. trading as Rain. The yield is Aave V3. Custody is pushed to the user. What Tuyo builds is an iOS and Android app plus an integration layer.

CREDIT Tuyo describes itself accurately and says so first: "Tuyo Inc. is a financial technology company. Tuyo is not a bank, cryptocurrency exchange, or asset custodian. Tuyo facilitates self-custody and offers a card issued by a third-party card issuer" (App Store listing, 2026-09-17). The press calls it a neobank. It is not one and it does not claim to be. That first-party honesty is rarer in this category than it should be.

1. Jorge Izquierdo co-founded Aragon in 2016, wrote the core protocol himself (940 commits to aragon/aragonOS against 135 for the next contributor), was CTO then CEO, and resigned on 2021-01-11 in protest over governance decisions, 32 months before Tuyo was founded and 34 months before the Aragon Association announced its dissolution. Alejandro Perezpayá joined as second co-founder in 2024-08, eleven months after founding, so this is not an Aragon reunion.

2. No Aragon money is traceable to Tuyo. THREE CHECKS, ZERO HITS The 29-page Patagon Management complaint (S.D.N.Y. 1:24-cv-08290, filed 2024-10-31), the most motivated document in existence for finding founder self-dealing and which did name a different venture (Nation3), contains zero occurrences of "Tuyo" and never names Izquierdo as a defendant. 529 transactions across izqui.eth, cuende.eth and tuyo.eth were cross-referenced against five known Aragon treasury addresses: zero hits. And the full unclaimed residual of 11,249.01 ETH reconciles exactly, to the penny, from the redemption contract through a Safe to the Aragon Foundation treasury on 2024-12-17 and 2024-12-19.

3. Tuyo could nonetheless have been seeded lawfully from personal money, and that is the likeliest answer. The 2017 sale terms allocated 15% of the ANT supply, about 6.48M ANT, to "the founders and early contributors," vesting over two years with a six month cliff and so fully vested around 2019. That pot was personal property years before Tuyo existed. The treasury itself was the Association's, held under a Swiss non-profit purpose, and could not lawfully have capitalised a private company. The per-person split is not public, so the size of Izquierdo's own pot is unknown.

REAL The wallet is genuine self-custody

A BIP-39 mnemonic on device (English word list, BIP-32, path m/44'/60'/0'/0/0, no passphrase, mnemonicToAccount from viem/accounts) derives a signer EOA that authorises a ZeroDev Kernel 3.2 smart account holding the assets. Both contracts are ZeroDev shared infrastructure live on Base, and user accounts resolve as ERC-7760 minimal proxies pointing at the Kernel implementation. The documented architecture is the deployed architecture.

Tuyo documents this in technical detail most of the category does not publish, including the warnings against itself: that reproducing an address does not prove authority, and that "The ejection route was simulated but has not been verified through a completed on-chain withdrawal in this guide." (help.tuyo.com/guides/understand-your-smart-account.html and recover-kernel-assets.html, 2026-09-17.)

CONTRADICTED The card layer is not self-custody, and the FAQ says it is

Each card user gets a CollateralAccount minimal clone. Reading the verified source (0x221677d88E507D21b4f7eC4D8D17e0f2FEd526af, Blockscout, 2026-09-17):

• collect(token, to, amount) is onlyCounterparty and needs no user signature. A Tuyo-controlled counterparty pulls funds unilaterally.
• release(...) is onlyDepositor but requires a valid counterparty signature, so the normal user withdrawal needs Tuyo to co-sign.
• The only unilateral exit is eject() then releaseEjected() after a cooldown read as 259,200 seconds, 3 days, in all 9,000 account-creation events. Ejection cancels every Tuyo card and cannot be undone, and Tuyo's own docs state the counterparty "can still collect during the cooldown, so the starting balance is not a guaranteed final payout."

Against that, tuyo.com/faq, same date: "Can I always take my money out? Yes... No lock-ups, no exceptions." For a card balance that is not accurate. There is a lock-up, there is an exception, the unilateral exit destroys the product, and the final amount is not guaranteed. The claim is true for the Kernel account and false for the collateral account, and the FAQ does not draw the distinction.

SHARPEST RISK One hot key governs all 9,000 accounts

CollateralAccountController.owner() returns 0xc87D11AE8c11a4c5a7BA3Ea48827bFFaecae1920. Confirmed today by direct eth_getCode: empty bytecode, so an EOA, not a multisig. That single key sets the counterparty and the collect whitelist for all 9,000 user collateral accounts, and owns both Earn vaults where _setFee permits a performance fee up to 100% of yield (principal is protected: emergencyRescue cannot touch the aToken). It is operationally hot: recent transactions include withdrawAsset on Rain's controller (2026-09-05, 2026-09-10) and addAdmin on Tuyo's Rain proxy (2026-09-15). Not disclosed anywhere Tuyo publishes.

CONTRADICTED Earn is one Aave wrapper, not a curated set

Marketing says "We curate for you the best yield strategies available in the market" and "The highest APYs on USDC, EURC and even BTC." What exists onchain is two proxies pointing at Aave's own ATokenVault (source header: @author Aave Protocol, "All Rights Reserved © AaveCo"). Tuyo wrote no strategy code. One protocol, not a curated set. No EURC vault and no BTC vault exist. The largest Earn user holds 20,000 EURC and 0.44 cbBTC as plain idle tokens, not vault shares. The ETH vault holds 0.18 WETH and is dead, which I confirmed today.

TWO REAL POSITIVES, UNADVERTISED Tuyo deployed Aave's audited vault rather than writing its own, so Earn-specific risk is about as low as this design gets. And the EIP-1967 admin slot on both vaults reads 0x...dEaD: the upgrade right has been burned and Tuyo cannot swap the implementation. For a self-custody product that is the right call.

Assurance, and known user-facing failures

No audit of any Tuyo-authored contract was found. No bug bounty. security.txt 404s on four hosts. The genuinely Tuyo-written code is the collateral system, which is also the part holding the card float and the part with no published assurance. It is short, roughly 120 lines, which lowers risk, but short is not audited.

VERIFIED NEGATIVE Tuyo holds nothing

This is now a verified negative rather than an absence of evidence. The FinCEN MSB register was swept across all 59 US jurisdictions, 32,462 registrant records parsed, returning 0 hits for "Tuyo", 0 for "Signify", 0 for "Izquierdo", 0 for "Perezpay" (msb.fincen.gov, 2026-09-17). Not on the NYDFS virtual currency list. No CASP authorisation. No state licence disclosure page, which every licensed transmitter must publish and which Bridge does publish in detail.

For a genuinely non-custodial wallet this is the expected and defensible answer. FinCEN's 2019 guidance treats an unhosted-wallet software provider that never controls user funds as outside the money transmitter definition, and Tuyo's architecture is built to sit exactly there. Terms 2.4: "The Service is a purely non-custodial application, meaning we do not ever have custody, possession, or control of your digital assets at any time."

Every permission is borrowed, from two suppliers, and one is now inside a competitor

Stripe closed its $1.1bn acquisition of Bridge on 2025-02-04. So Tuyo's fiat rails sit inside a company building competing consumer stablecoin products, with no commercial reason to prioritise an 11-person customer, and which is simultaneously the Program Manager for Phantom Cash. If Bridge goes, Tuyo is a wallet with no way in or out. If Rain goes, there is no card, and the card is the entire proposition. Replacement is a six to twelve month integration during which the product is dead.

Tuyo has no fallback permission and I found no evidence of a second provider or of a licence application in progress. This is standard fintech-on-rails risk and not disqualifying by itself. Most of the category runs this way. But the usual mitigation, either your own licence stack in progress or a redundant provider, is absent.

Footprint is narrower than the marketing

The pitch is "Receive payments like a local anywhere in the world." 27 countries are blocked platform-wide (no wallet, no Earn, no ramps, no card, no account numbers) including the United Kingdom, India, Nigeria, Vietnam, Turkey, Bangladesh, Nepal and Ukraine, which removes most high-volume inbound remittance corridors. The card is unavailable to residents of 19 US states plus DC and 5 territories, and New York additionally loses fiat at the Bridge layer.

Rain America is federally MSB-registered in every single state where the card is blocked, so the exclusions have nothing to do with federal registration. They are state licensing or issuing-bank policy, set by a private schedule nobody publishes. The card is not available in Delaware, the state whose law governs the Terms and most likely the state of incorporation.

Risky feature 1: Buy Now, Pay Maybe

Terms section 3.2 (revised 2026-07-28) define it as "a discretionary processing-discount feature pursuant to which Tuyo may, in its sole and absolute discretion, elect not to debit some or all of the amount that would otherwise be debited", with the merchant paid in full and the difference "absorbed by Tuyo." The same section asserts BNPM "is not a sweepstakes, lottery, prize, contest, raffle, game of chance, or rewards program."

The FAQ says the opposite: "some of your card purchases are randomly free."

The US lottery test needs prize, chance and consideration. Tuyo's drafting attacks two of three. On consideration it is strong: no opt-in, no enrollment, no minimum spend, and nothing a user can do to change the odds. That is better than a bolt-on alternative method of entry because there is nothing to enter. On prize, "a discount, not a transfer of value" is orthodox. On chance it simply asserts the conclusion and the marketing contradicts it.

DRAFTING TELL The same Terms added an Invite Program in the 2026-07-28 revision whose section 3.16.7 expressly states its vouchers "are earned deterministically upon satisfaction of Tuyo's applicable activation criteria and are not distributed by chance." The BNPM section has no equivalent sentence. In a document this carefully worded, negating chance where you can and declining to where you cannot is not an oversight.

No regulator has acted against a comparable mechanic. Nothing in CourtListener, SEC releases or press. That is novelty, not a clean bill of health: no safe harbour to point at and no adverse ruling to underwrite against. Context, not precedent: Congress had to pass the American Savings Promotion Act of 2014 to permit prize-linked savings, and only for banks and credit unions, which tells you which way the default runs. Five states legislated against sweepstakes mechanics in 2025, and New York's statute reaches supporting service providers including payment processors.

My read on the near-term risk is not gambling law. It is unfair and deceptive practices. Put "Your next purchase may be free" and "randomly free" next to "not a sweepstakes, lottery, prize, contest, raffle, game of chance" and a state AG sees a consumer sold a chance outcome by the marketing while the contract denies the chance outcome exists. Cheaper to bring, and it does not require anyone to decide whether BNPM is gambling.

Risky feature 2: retail yield up to 11 percent

Tuyo is structurally better placed than the enforcement comparators, and it is worth being precise about why.

All four took the asset, pooled it, deployed it at their own discretion and promised a rate, so the user held a claim against the platform and nothing else. Tuyo keeps the key with the user, routes to third-party vaults on a public chain, promises nothing, and discloses that returns may be "lower, zero, or negative." Those disclosures are materially better drafted than anything BlockFi or Celsius published.

What weakens it is not small. "We curate for you the best yield strategies available in the market" is managerial effort in Tuyo's own words, and that is the efforts-of-others limb Howey turns on. The 10% performance fee makes Tuyo an economic participant in the user's return rather than a neutral pipe. And there is real tension between "we curate for you and take 10%" and "we owe you no fiduciary duties and are not a party to these transactions." Both cannot be true in the way a user would understand them.

One EU issue for counsel, flagged as an issue and not a conclusion. EURC is an e-money token under MiCA, MiCA restricts granting interest on e-money tokens, and Tuyo advertises "the highest APYs on USDC, EURC and even BTC" to a base including EEA residents, while holding no authorisation of its own to be supervised under.

3.5x GAP The 11 percent does not survive a check

Reading convertToAssets on tUSDCb at historical blocks gives realized net yield to depositors, after the 10% fee, measured from live chain state:

Every window lands near 3 percent. The advertised figure is "up to 11% APY" and "the highest APYs on USDC." Two fair caveats: "up to" is doing work, and the headline may blend a TUYOs points value that has no price because the token does not exist yet. But whatever 11 percent describes, it is not what depositors in the only material Tuyo vault have received, and the acquisition promise is therefore not deliverable from the actual strategy.

CREDIT The 10% fee itself is exactly what the contract enforces. getFee() returns 100000000000000000 against a SCALE of 1e18. I re-confirmed this today: precisely 10.0%. That disclosure is verifiable and correct.

THE PATTERN Three marketing-versus-contract contradictions in one sweep

Add a privacy policy last updated 2024-07-17, two years stale, through the launch of the card, the launch of Earn, EEA and Mexican fiat rails and the Invite Program. No GDPR controller identity, no legal bases, no transfer mechanism, no retention periods, roughly 2,800 characters, at a company processing KYC data for EEA residents. The Terms were revised seven weeks ago. The privacy policy was not.

Each is fixable in an afternoon. Together they say the compliance function is not keeping pace with the product, at a company whose entire legal position depends on those documents being exactly right. The engineer's documentation on help.tuyo.com is honest and specific. The marketer's copy is not. Nobody is reconciling them.

Which of them do the exact Tuyo bundle

The verdict. Exactly one company shipped Tuyo's full bundle to consumers before Tuyo, and it concluded the consumer business was not the business. Three companies are assembling the same bundle now at 100x to 300x Tuyo's scale. Tuyo owns none of its four legs. The account numbers belong to Bridge, the card to Rain, the yield to Aave, and self-custody is a posture any wallet can adopt in a sprint, not an asset.

The graveyard, and the recurring cause of death

1. Partner failure, not competition. Visa terminated WaveCrest on 2018-01-04 and Wirex's roughly 500,000 cards went dead in one second ("All the cards were shut down in one second", co-founder Dmitry Lazarichev). Wirex today has 1,407 lifetime iOS ratings and has not raised since 2022-02-16. Juno survived losing Wyre in 2023, then the Synapse and Evolve failure killed it: all crypto services retired 2025-09-30, and Harmonic now shows OnJuno at $0 funding, 249 web traffic and 100 Twitter followers. Binance ended EEA Visa card services 2023-12-20 as Contis and Solaris stepped back. Not one of these lost a product fight. Tuyo rents two regulated layers.
2. Card economics never covered the promised rewards. Crypto.com cut CRO card rewards from a 1-8% range to 0-2% in May 2022, CRO fell 11-17% on the day, and the CEO partly reversed within 24 hours after backlash. Fold listed on Nasdaq at around $10 in Feb 2025 and the stock plunged to $3.32 on announcing its bitcoin rewards credit card. Public markets do not believe card rewards are a business.
3. Yield is a securities question. Linus stopped selling. Celsius raised $900.3M and BlockFi raised $1.266B at a $4.75B valuation; BlockFi now has 9,516 monthly web visits.
4. Self-custody users do not spend. The clean natural experiment: Robinhood's main app has 4,840,585 US iOS ratings and Robinhood Wallet, its self-custodial product, has 6,701. That is 0.14 percent. Same brand, same marketing machine, same install base to cross-sell.
5. Mid-tier wallets get squeezed from both ends. Leap Wallet's epitaph, 2026-05-28: non-custodial wallets "operate on thin, flow-driven revenue streams, which makes mid-tier players particularly vulnerable as activity and liquidity concentrate."

The stalled middle, which is the tier Tuyo is in

Not one self-custodial consumer wallet in this middle tier has raised a priced up-round in the last four years.

The counter-case is real and narrow. KAST raised $90M at $600M with 1,172 ratings and RedotPay raised $194M with 16,408. Capital is available in this category at valuations that do not track users. If the thesis is "the category is funded aggressively and Izquierdo can catch a round on reputation," that is a defensible read of the market. It is a financing thesis, not a business thesis, and both comparables are custodial, which is the one design decision Tuyo has ruled out.

THE FINDING The fee schedule excludes Tuyo's own core use case

From the FAQ, quoted: "Bridging tokens and trading between stablecoins are exempt from the volume-based fee", and there is "no fee for USD transactions." The flagship flow is on-ramp to USDC, earn on USDC, spend USDC. Under Tuyo's own published schedule that flow generates exactly one revenue line: the 10% Earn performance fee. The 25 bps trading fee only bites when a user trades into volatile assets, which is wallet behaviour, not money-app behaviour.

Earn is the only scalable line, and it needs $100M to produce $1M

Displayed APYs are already net, so if the displayed rate is r the take is r × 0.1111 of balance per year. At the 11% headline that is 1.22%; at a 9% displayed rate it is 1.00%, the working figure.

Tuyo's Earn book today is $375,480, read from the contract by me. At a 1% take that is roughly $3,800 a year, and the 1,302 USDC of accrued uncollected fees corroborates the order of magnitude.

The card, and why Buy Now Pay Maybe is a bind

Interchange. Federal Reserve Reg II data for 2024: exempt issuers (under $10B assets) average 1.21% of value; covered issuers 0.47%. In the EU, Regulation (EU) 2015/751 Article 3 caps consumer debit interchange at 0.20% by law. That pool is split among the issuing bank, Rain, the processor and Tuyo. Assuming Rain and the issuer keep 60%, which is my assumption and is unverified, Tuyo nets about 0.48% on US spend and 0.08% on EU spend. At Tuyo's observed 30% US mix that blends to about 0.20% of card spend. If Tuyo receives no interchange at all, which is possible under a plain platform agreement, this line is zero.

Add FX at 0.54% effective (0.9% spread on an assumed 60% non-USD share) and gross take on card spend is 0.74%.

So Buy Now, Pay Maybe breaks even at a forgiveness rate of 0.74% of spend, which is one transaction in 135. Both settings of that parameter are bad. At one in 135 the giveaway is statistically invisible: a user tapping a few times a week goes months without a hit and "Imagine buying coffee and not paying" reads as false. At a rate high enough to be the product, say one in 20, the cost is 5% of spend against a 0.74% take and the card loses 6.8 times what it earns. Tuyo discloses neither the odds nor any cap. The entire card contribution margin hinges on one undisclosed parameter.

Per-user, and what the targets require

From card spend alone, at 0.20% blended net interchange, $1M of revenue needs $500M of annual card spend, which is nearly four times the entire Gnosis Pay network's $131M across all of 2025. $20M needs $10B, roughly 76 times that network.

Implied revenue today at 1,000 to 3,000 engaged users is $39,000 to $117,000 a year. Given 428 lifetime iOS ratings and roughly 800 collateral accounts holding $100 or more, the low end is likelier. This is a pre-revenue company by any practical definition, and the gap is a user-count gap, not a pricing gap. Doubling every fee does not change the answer.

Every competitor that makes money on a card charges for it

Five levers exist in this category: charge a subscription, fund rewards with a token, sell infrastructure, monetise swap flow, or already own the wallet. Tuyo has picked none of them and runs the highest-cost promotion in the category.

The curve is points farming, not a business curve

Growth is spiky, not compounding. Three spike months and troughs between them. That is the signature of incentive-driven acquisition, not organic pull. Two thirds of card accounts have a nonzero balance and half of those have less than fifty cents. These are accounts that were opened once and abandoned.

VERIFIED There is no round

Two independent databases, one regulator and the alleged lead investor's own disclosure all say the same thing. The one named investor claim does not survive contact.

THE ANSWER The financing instrument is a token

From tuyo.com/faq and tuyo.com/rewards, both retrieved 2026-09-17:

• "When is Tuyo launching a token? We expect to launch a token in 2026."
• "Is there going to be an airdrop? At token launch, Tuyo will allocate at least 20% of its supply to users and community."
• "Get TUYOs. Own Tuyo. ... Users of Tuyo are not just customers, but owners."
• Season 0 Pioneers to 2025-11-01. Season 1 Early Adopters. Season 2 FOMO. Season 3 "Stacking Season" running from 2026-01-22 until TGE (2026).

There is no contradiction with "TUYOs are not a cryptocurrency token." Both are true: TUYOs are non-transferable points today and the accounting ledger for a token allocation tomorrow.

This reframes the traction. The November 2025 spike of 1,333 against 313 the month before lands exactly on the start of Season 1. Season 3 has run for eight months with no TGE. A material share of the 9,000 accounts and of the $375,480 in Earn is airdrop farming, which is consistent with a $0.43 median collateral balance and 15 addresses holding 91% of Earn.

Ranked explanations for the $3M to $5M plausibly spent

1. Founder capital. Most likely. Fits every observable: no Form D, no Harmonic round, no investor claiming the company, no announcement, and no pressure to raise while posting US market salaries. Izquierdo's 2017 founder allocation vested around 2019 and was lawfully his. No direct evidence of his holdings exists and none is invented here.
2. An undisclosed private round. Plausible, uncorroborated. A Reg S structure sold only to non-US persons, or a token warrant rather than equity, files no Form D and is consistent with two Spanish-speaking founders and a Delaware entity.
3. A token sale already done. Ruled out on the public record. No TUYO token trades on any chain (DexScreener, all chains, 2026-09-17). A private SAFT would not be visible and cannot be excluded.
4. Revenue. Real but not covering burn. Accrued uncollected Earn fees across both vaults are 1,302 USDC and 0.0022 WETH. That is the entire uncollected Earn revenue, and it sizes the business honestly.

BRAND DAMAGE ALREADY ACCRUING 24 impostor TUYO tokens on Base

Tuyo is unambiguous: "TUYOs are not tradable nor transferable... Anyone who sells TUYOs might be trying to scam you." Searching Base returns roughly 24 tokens carrying the TUYO ticker or the Tuyo name. The three largest hold 5,676, 3,427 and 3,353 holders from three mutually unrelated creator addresses, none matching Tuyo's deployer or admin. Over 12,000 holder positions sit across the five largest fakes. Three separate Base tokens are literally named with Tuyo's own truncated referral message, so someone is farming users who paste the referral text.

Tuyo publishes no canonical contract address list anywhere, and it has announced a TGE. That is the single cheapest thing it could do about this, and it has not done it. The prior Daxos read of "wrong or premature" resolves to premature, and being actively front-run.

DAXOS FIT 2.0 / 10. No cheque. There is nothing to buy.

The binding fact is mechanical, not judgmental. Harmonic records currently_raising null and 0 rounds. There is no Form D. No investor claims the company and the one named claim was refuted against the alleged investor's own portfolio page. A fund cannot underwrite a company it cannot buy, and Daxos writes $100k to $500k into rounds that exist.

Worse for an equity buyer, the capital formation instrument appears to be a token. The company has publicly committed to a TGE in 2026, structured its entire rewards programme as numbered pre-TGE seasons counting down to it, told users they are owners, and promised at least 20% of supply to users and community. If that is the real capital event, an equity cheque has a structural problem: value accrues to token holders, and 20% of supply is committed to users before any investor allocation exists. Whether equity participates pro rata in the token is not discoverable from outside and it determines whether Tuyo is investable by Daxos at all.

Three realistic paths, in descending plausibility: a pre-TGE token instrument or SAFT, which is a different instrument with different risk, jurisdiction exposure and downside than equity; buying on market after the TGE, which is trading and carries no allocation advantage; or a priced equity round that does not currently exist.

Price, if a priced seed ever appeared

No equity price is defensible without first seeing the token and equity split. The honest ceiling is single-digit millions post. The anchor is Bleap, which Harmonic records at $8,300,000 raised across 3 rounds at SEED with 12x Tuyo's web traffic, a comparable app base, a real SEO acquisition engine and its own MiCA licence. Tuyo has less of everything except the founder, and rents the layer Bleap owns.

So not above a $6M to $8M post, and only with a pro-rata token warrant attached. For context on why a higher number cannot be reasoned into: KAST cleared $90M at a $600M valuation on 1,172 app ratings, so the category price signal is loose enough that a founder could plausibly ask far more. That is a reason to be more careful, not less.

VERDICT: MEET. One hour, relationship and information, explicitly not a cheque.

The justification is narrow and I want it stated so nobody misreads it. No answer to any question below moves Base above roughly 3.5, because the market, the model and the measured traction are unchanged by anything Izquierdo can say. The meeting is worth taking because five specific factual questions are answerable only by him, because the founder is reachable at j@tuyo.com (Harmonic primary_email, person 132850), and because Daxos's edge in this category is founder relationships rather than allocation. If the answer to question 1 is "equity does not participate in the token," this converts to PASS the same day.

1. What is the split between equity and token, has any token supply already been allocated, to whom, at what price, under what jurisdiction and lockup, and does equity carry a pro-rata token warrant?
2. Is BNPM genuinely randomised, how is the state exclusion list determined, and why is it not published?
3. 0xc87D11AE8c11a4c5a7BA3Ea48827bFFaecae1920 is a single EOA that owns the collateral controller and both Earn vaults, can set the counterparty and the collect whitelist across all 9,000 user accounts, and can set the performance fee to 100%. What is the key custody arrangement and is a multisig planned?
4. The FAQ says "No lock-ups, no exceptions" and CollateralAccount has a 3 day cooldown, a counterparty co-signature on normal release and a counterparty collect needing no user signature. Which describes a card balance, and will the FAQ be corrected? Relatedly, what does "up to 11% APY" measure when the realized net figure is 3.07% over 90 days?
5. Has any capital been raised to date in any form from anyone, and if so why is there no Form D? Plus the cheap one: what is the state of incorporation, and can you send the good-standing certificate?

A meeting is warranted now. A process requires all of the first four. A cheque requires all six.

Additional triggers that would reopen this file on their own: a Tuyo-authored contract audit published; a second fiat or issuing partner signed, which removes the single points of failure; a licence application filed anywhere in Tuyo's own name; or Rain or Bridge terminating, which would end the file rather than reopen it.

What would close it permanently: equity confirmed not to participate in the token; a raise above $15M, which trips the house hard filter; or any enforcement action on BNPM or Earn.

• State of incorporation. Not confirmed from any registry. California SOS bizfile returned HTTP 429 blocking the droplet IP by name. Delaware ICIS returned an error and prohibits automated queries. Florida Sunbiz returned 403 on two network paths. Delaware is inferred from the governing-law clause, a 302 phone number, the CSC agent address on the Play developer record and one press report. Five minutes on a residential connection or a $20 agent search would settle it, and it should be settled before any cheque.
• Registered agent, directors, officers beyond the two founders, corporate good standing. No registry access, so none of it.
• Whether any Spanish or other non-US entity exists. None disclosed, none found. Libreborme returned 403, CNMV returned 400 and 403. Unchecked, not a confirmed negative. This matters because an undisclosed EU establishment changes the MiCA analysis.
• Tuyo's actual funding. Harmonic records zero and flags EXISTS_BUT_UNDISCLOSED, which is a model inference. No Form D exists. House rules bar taking a number from the web and none was taken.
• Revenue, deposits, card spend volume, funded accounts, CAC. None published, no credible third-party estimate. Everything in the business-model section is a model, not a measurement.
• Buy Now, Pay Maybe odds and caps. Not disclosed anywhere. The entire card contribution margin depends on this one number.
• Whether BNPM is genuinely randomised. The FAQ says randomly, the Terms decline to negate chance while expressly negating it for the Invite Program. Strong inference, not confirmation. Only Tuyo can answer it.
• Whether Tuyo receives any interchange and on what split. The 60/40 assumption to Rain and the issuer is mine and is unverified. If the real answer is zero, the card is a pure cost centre.
• The bank that actually issues the Tuyo card. Not named by Tuyo anywhere. Rain names "Third National" and "SSB, Member FDIC" and neither resolves to a single institution in FDIC BankFind.
• Rain's state money transmitter licences. Rain publishes no list at all, unlike Bridge which publishes 32 jurisdictions with numbers. The card's 20-jurisdiction US exclusion is almost certainly explained by it, but the schedule is private.
• CFPB complaints. API returned 403, the legacy Socrata dataset is retired, the UI needs JavaScript. Not checked for Tuyo, Bridge or Rain. A genuine gap, partly mitigated because complaints about a card would log against the issuing bank Tuyo does not name.
• FTC and CNMV. Press only and blocked respectively. Partial checks. NMLS Consumer Access returned HTTP 403 on every attempt; the inference is safe because Tuyo publishes no licence disclosures and a licensed transmitter must, but NMLS was not queried directly.
• iOS version history. Apple serves it from an authenticated endpoint and the bearer token is not in the page. Release cadence is inferred from two data points, not established.
• The March 2026 Earn withdrawal freeze. One uncorroborated App Store review. The status page retains only 90 days. Cannot confirm or refute, and it matters more than its single-review weight because Aave's ATokenVault has no owner function that blocks redemption, which would place any such block in Tuyo's app layer.
• Seven contracts Tuyo deployed on 2025-10-01 and 2026-09-15 do not respond to name(), symbol(), asset(), totalAssets() or owner() and have unverified source. Purpose unknown. Two were deployed the day before the latest app release.
• Extrapolation precision. Card collateral of about $490k rests on a 750-of-9,000 sample with zero failed reads and is reasonably tight. Idle wallet USDC of about $242k rests on 15 nonzero observations and is an order of magnitude only. The Earn and vault figures are exact contract reads, re-confirmed today.
• Earn holder coverage. 95 live holders covering 92.1% of supply. About 7.9% sits with holders not enumerated, so concentration could only get worse, not better.
• Whether the $271,239 in Tuyo's Rain collateral proxy is corporate capital or aggregated user funds. Chain data does not settle the accounting. Excluded from the $1.1M user-asset figure rather than folded in.
• Five of eleven employees are unnamed. Harmonic returns 6 people records for headcount 11. The Apify LinkedIn profile scraper failed twice with HTTP 400 and was stopped at the two-error limit, so the 3-profile cap was never spent.
• Completeness of the Aragon onchain negative. Five Aragon addresses against three ENS-resolved addresses, Ethereum mainnet only, on a capped result set. ENS ownership is not proof of identity. Clean but not proof of absence. Any settlement in Patagon v. Aragon is not on the public docket; a Rule 41(a)(1) voluntary dismissal discloses no terms and PACER was not paid for.
• Revolut. Harmonic returned an enrichment pointer with no entity payload on three attempts, so Revolut is unquantified rather than estimated from the web.

Distribution placement. At 4.50 company, Tuyo sits in the band the house tiering heuristic calls "pass, or micro check on relationship." The Daxos-fit number of 2.0 is the one that governs the action, and it says no cheque.