Inside Noxa:
the multichain launch stack behind WALLET
Noxa rehearsed this stack on MegaETH, then brought it to Robinhood Chain before public mainnet. It launched 4663, administered WALLET's fee path, and later shut revenue to zero. New tracing resolves every treasury bridge, corrects a false funder lead, and places WALLET among just 11 zero-buy launches in a 151-launch cohort.
What the chain can answer about Noxa, and what still needs proof.
The community theory contains several distinct claims. Testing them separately produces a stronger result than treating Noxa as either an innocent tool or a confirmed Robinhood operation.
Did Noxa rehearse its tooling on other chains, arrive early on Robinhood Chain, protect liquidity, keep the system running, and then leave without extracting value?
#41841 remains in the verified non-proxy LaunchLocker. Its reviewed code exposes no liquidity-decrease, NFT-transfer, arbitrary-call, or withdrawal path.f70d reveal Noxa's backer?f70d is Relay's published EVM solver address. The address dcef deposited its own 0.1 ETH into Relay on Ethereum and received the solver fill on Robinhood Chain. Its earlier source was an unshield from RAILGUN, where public attribution ends.Mercury ui_multiplier and ui_quantity integration. Mercury has not been publicly attributed, but the comment makes the test more product-specific than the transaction sequence alone showed.What this means for WALLET
Noxa accounts for the mature launch machinery and its early availability. aa33's separate provenance remains: funding the future OpenAI stock-token demo wallet, correcting an ERC-8056 display contract, operating during restricted mainnet, and receiving funding from F718. The first WALLET investigation documents those links.
Noxa had already built and operated this system on MegaETH.
Official Noxa records identify the earlier deployment. Runtime comparison proves the Robinhood contracts reused the same launch architecture.
13 November 2025
- LaunchFactory0xAc3039...45B9a
- LaunchLocker0x93e797...14733
- Earliest visible public activity9 Feb 2026
16 June 2026
- LaunchFactory0xD9eC2d...6FCcB
- LaunchLocker0x7F03ef...acD85
- Before public mainnet15 days
5124fdbb984a6eb362776232f13c197e3c1d36e1abbb09c7cbeb6de65bf7e051d70fbe441441da5cfe05588f9ae855e2342b39e1ee67362775fac41c7ab07a4ceth_getCode was called against both chains. The 4,823-byte locker runtimes are identical. Each 22,811-byte factory contains its chain's locker address five times; replacing those five 20-byte values produces identical runtime and the normalized hash above.Noxa's own stated model: it deploys a complete DeFi stack on new blockchains before others. MegaETH and Robinhood Chain show that model in code. This supports a rehearsed early-chain strategy. It does not show how Noxa obtained Robinhood Chain access or establish Robinhood authorization.
The same launch system moves through three distinct phases.
MegaETH rehearsal leads into restricted Robinhood deployment, then public operation, WALLET, and an onchain revenue wind-down.
MegaETH proves the tooling existed first.
The factory and locker were deployed together, then used publicly months before the Robinhood Chain sequence began.
The same creator, 0xC31245...1be03, deploys both contracts one second apart.
Launch and fee-collection activity shows the architecture working in production.
The system arrives before public access.
Funding, contract deployment, and Noxa's own 4663 launch form a clear pre-mainnet operating sequence.
After receiving 0.109725 WETH from RAILGUN on Ethereum, dcef deposits 0.1 ETH into Relay. Relay solver f70d fills the same address on Robinhood Chain; dcef trades, then sends 0.005 ETH to dev.noxa.eth.
Noxa redeploys LaunchLocker and LaunchFactory 15 days before the public-mainnet announcement.
dev.noxa.eth calls launchToken and records treasury.noxa.eth as deployer. The token's name and description do not prove Robinhood authorization.
Fees, WALLET, shutdown, and wind-down.
Public access expands the system. WALLET follows the standard launch path, then Noxa closes launches and removes its revenue shares.
Noxa adds configurable token-treasury shares, protocol shares, and per-token overrides.
The public factory receives a 0.0005 ETH fee and creates position #41841 in the standard locker.
Noxa's public explanation cites bot spam and infrastructure pressure.
Noxa removes the token-treasury share, cuts the protocol share to 25%, then sets it to zero.
Why build the system, open it to the public, and then walk away?
The deployment purpose, the public shutdown explanation, and the onchain wind-down are three separate pieces of evidence. Together they support a bounded answer, but not a proven hidden motive.
Noxa appears to have arrived to establish early-chain launch infrastructure, operated it through the public opening, and then executed a deliberate shutdown when its stated operating limits were reached.
CoinDesk reported nearly $12 million in cumulative platform fees before the shutdown. That figure is not the same as Noxa's retained profit, but the admin sequence still surrendered future protocol revenue. A temporary bootstrap assignment, planned handoff, risk decision, or non-financial mandate therefore remains a serious alternative to a conventional profit-maximizing business.
Noxa came to deploy launch infrastructure.
Noxa described its model as bringing a complete DeFi stack to new blockchains before other operators. The matching MegaETH deployment, pre-public Robinhood contracts, and Noxa's direct launch of 4663 fit that stated strategy.
The exit was administered, not an accidental disappearance.
Noxa disabled new launches, removed the token-treasury share, reduced the protocol share to 25%, and then set it to zero. Those ordered admin actions show an operator winding down live economics after stopping the factory.
Noxa publicly blamed bot spam and infrastructure pressure.
The chain confirms the shutdown actions, but it cannot show whether operating pressure was the sole cause. A planned temporary deployment, commercial handoff, risk decision, or completed bootstrap mandate remains possible and unproven.
The sequence fits an experienced early-chain operator deploying proven infrastructure, bootstrapping public activity, and conducting an intentional wind-down after an operating or exit condition was reached. For WALLET, the net effect is positive: no Noxa-specific allocation or fee override was found, no wind-down action targeted WALLET holders, and the reviewed base LP remains locked. The chain still cannot reveal whether Noxa acted independently, under a commercial arrangement, or as temporary infrastructure for another party.
Any wallet could call LaunchFactory; Noxa controlled protocol fee settings and later overrides.
The admin, recorded token deployer, and public factory caller held different powers. Collapsing them into one role overstates Noxa's authority.
The first analysed receipt shows why Noxa's treasury held WALLET.
Position #41841 generated both assets. The locker collected them and passed them into the FeeManager route then in force.
#41841
58.811m WALLET
3.169356 WETH
LaunchLocker
Collects fees; reviewed code exposes no liquidity-decrease or NFT-transfer path.
FeeManager
Applies token-side and WETH-side receiver shares.
58,810,972.250227564802082944 WALLET
11,762,194.450045512960416588 WALLET to treasury; 47,048,777.800182051841666356 to 0xdead.
3.169356103741596815 WETH
2.060081467432037929 WETH to treasury; 1.109274636309558886 WETH to aa33.
WALLET followed the default Noxa contract path; aa33 remained exceptional.
The receipt comparison currently covers WALLET and two sampled launches, WISHBONE and IF. That establishes matching examples, not a population-wide baseline.
| Test | WALLET | Two sampled comparator launches | Evidence grade |
|---|---|---|---|
| Factory entry | Public launchToken; 0.0005 ETH fee | Same function and fee | Strong match |
| Creator allocation | No launch allocation transfer to aa33 | Not yet cohort-tested | Open baseline |
| Initial buy | initialBuyAmount = 0 | Not yet cohort-tested | Open baseline |
| LP architecture | One-sided V3 position #41841 | Same reviewed locker generation | Supported |
| Fee override | No direct override located | Default unless admin calls setOverride | Not found in tested set |
| Token fee split | 20% treasury / 80% burn | Same in WISHBONE and IF receipts | Two-receipt match |
| WETH fee split | 65% treasury / 35% deployer | Same in WISHBONE and IF receipts | Two-receipt match |
| Pre-launch top-level contact | None found with known admin or treasury | Not required by public factory | Not found in tested set |
Among aa33's 71 indexed top-level outbound transactions on Robinhood Chain, the WALLET launch is the only direct call to the known Noxa contract and admin set. This test cannot exclude offchain communication, intermediaries, cross-chain contact, account-abstraction routing, or an unidentified Noxa wallet.
All 14 bridge deposits are resolved: 33 ETH returned and 148.1 ETH completed cross-chain.
The treasury accumulated protocol fees, traded through routers, managed liquidity, and sent capital across chains. The new route-level reconciliation distinguishes bridge submissions, refunds, destination assets, and the point where public tracing ends.
token contracts
Blockscout snapshot includes protocol residue, dust, spam, and balances with negligible value. Balance count is not an investment count.
treasury.noxa.eth
115 signed outbound transactions reviewed; 113 succeeded. Only 52 successful transactions carried positive native value.
The successful native-value total remains 237.372254977414630919 ETH after excluding two reverted transactions. The 181.1 ETH bridge line is gross submitted principal, not completed outflow: a 3 ETH request and a 30 ETH request were refunded to the Robinhood treasury. The other 148.1 ETH completed across Ethereum and HyperEVM, sometimes with a destination swap. Delivered amounts were lower where bridge and swap costs applied.
The follow-up makes the WALLET thesis stronger and the Noxa story clearer.
The strongest new signal is comparative: WALLET launched with no creator buy in a cohort where 140 of 151 launches included one. Nothing in the follow-up shows Noxa-specific control, supply privilege, or an adverse action against WALLET.
The apparent funder became an infrastructure wallet, every bridge route gained a destination, and WALLET emerged as an unusually clean launch inside a high-volume public cohort.
More evidence of a deliberate clean launch. No new evidence of creator extraction or adverse protocol control.
The apparent funder was Relay infrastructure, not a Noxa-linked backer.
Identity neutralRelayAdapt / Ethereum0.109725 WETH exits the private pool. The depositor before this point is not publicly attributable.
0xdCef...8C607dcef receives the unshielded value, then deposits 0.1 ETH into Relay.
0x4cD0...8BC31Relay accepts dcef's origin deposit. Its published solver performs the destination fill.
solver: 0xf70d...dbEFThe same dcef address trades 0.058 ETH through SwapRouter02.
0x7E03...242Bdcef sends 0.005 ETH to Noxa's named development wallet.
0xf70d...dbEF is Relay's official EVM solver address, not an identifiable Noxa backer. The trace establishes a privacy-aware funding route into Noxa, while the source identity remains hidden behind RAILGUN.
The treasury managed 148.1 ETH of successful cross-chain principal.
Professional-operator signalETH across 14 deposits to Relay and LI.FI.
ETH across two completed refunds.
ETH routed cross-chain before bridge and swap costs.
0x92b3...A68a on Ethereum. Net delivered: 74.591435 ETH.treasury.noxa.eth on Ethereum as ETH, WBTC, and XAUt.0x92b3...A68aThe six resolved routes delivered 74.591435 ETH to this address.
The receiver later shielded 39 ETH and 35 ETH. It separately shielded 19,950 USDT from additional inflows. Public attribution stops at the private pool without a viewing key.
WALLET was one of 11 zero-creator-buy launches in its immediate 151-launch cohort.
Direct evidence supportThe 75 successful launchToken calls before WALLET, WALLET itself, and the next 75 calls. Window: 10 July, 00:51 to 01:13 UTC.
The public factory path was ordinary. The launch choices were not. A zero creator buy avoided creating a creator position through the launch transaction, and the official Robinhood Wallet URL was unique in this 151-launch window. Users could enter any metadata, so the URL remains circumstantial evidence rather than attribution proof.
aa33's stock-token test named an integration target, and Robinhood Wallet now executes onchain perps.
Product roadmap contextaa33 deploys TestERC8056Token. Its verified source says it was built for “testing Mercury ui_multiplier / ui_quantity integration on Base.”
The same account launches WALLET with no creator initial buy and links its metadata to Robinhood's official wallet page.
Robinhood Wallet now places perpetuals trades through Lighter. USDG moves on Robinhood Chain into a Lighter relayer contract and users sign each action in the wallet.
The verified test used Solidity 0.8.24, optimizer 200, and EVM Paris. Those settings are common and do not identify a developer. The source-level Mercury reference is more specific, but Mercury has not been tied to Robinhood through a primary public source.
The Lighter integration is confirmed by Robinhood. No primary source located in this review supports a Robinhood-dYdX/Arcus partnership or a 50/50 Lighter revenue split. Those claims are excluded from the evidence chain.
Which findings support the clean-launch case, and which only describe Noxa?
Evidence gradedOnly 11 of 151 surrounding launches had no creator initial purchase. This strengthens the clean-launch and community-asset thesis.
The same deployer moved from a named stock-token UI integration test to WALLET. Robinhood's live Lighter integration confirms that its wallet is an onchain financial-execution surface. Neither fact proves WALLET token utility or aa33's employer.
Reused contracts, pre-public deployment, disciplined administration, cross-chain portfolio construction, and a staged wind-down support a professional-operator reading. They do not identify Robinhood as the principal.
Correcting the false lead removes an invalid Noxa-backer inference. The replacement path still shows privacy-aware funding into Noxa, but the pre-RAILGUN identity is unknown.
This weakens claims that Noxa never pursued value or kept all routes transparent. It does not show WALLET dumping, personal extraction, or any action against WALLET holders.
The evidence isolates Noxa's role and leaves aa33's proximity signals intact.
| Observed evidence | Supported interpretation | Still unproven |
|---|---|---|
| MegaETH locker runtime is identical; normalized factory runtime matches. | Noxa had rehearsed, reusable multichain launch tooling. | Robinhood commissioned or authorized Noxa. |
| Noxa deployed on Robinhood Chain 15 days before public mainnet and launched 4663 itself. | Noxa was an active early-chain operator. | Noxa was owned, hired, or directed by Robinhood. |
| In a 151-launch window, 140 launches included a creator initial buy; WALLET was among the 11 that did not. | WALLET used the public factory while avoiding the cohort's normal creator-positioning pattern. This strengthens the clean-launch thesis. | Robinhood directed the launch or supplied the metadata. |
| WALLET fees reached Noxa treasury through FeeManager. | Noxa administered and benefited from the default fee mechanism then in force. | Noxa bought, selected, endorsed, or held WALLET as a conviction position. |
f70d is Relay's published EVM solver. dcef self-deposited on Ethereum after receiving value from RAILGUN. | The apparent Noxa funder was bridge infrastructure; the real source trail terminates at a privacy pool. | The person or organization that controlled the pre-RAILGUN funds. |
| Of 181.1 ETH submitted to bridges, 33 ETH was refunded and 148.1 ETH completed. A receiver later shielded 74 ETH and 19,950 USDT through RAILGUN. | Noxa treasury capital was actively diversified, and part of the clustered flow crossed a public attribution boundary. | Personal extraction, a token sale, a Robinhood link, or the final private-pool beneficiary. |
| aa33 has separate demo-wallet, ERC-8056, restricted-mainnet, and F718 links. | WALLET's strongest proximity signals remain outside Noxa. | Robinhood attribution, employment, or formal mandate. |
The follow-up makes WALLET harder to dismiss.
The Noxa investigation adds context without damaging the clean-launch case.
Noxa had already deployed the same architecture on MegaETH and specialized in arriving early on new chains. On Robinhood Chain, it controlled the reviewed launch, LP-custody, and fee infrastructure, launched 4663 directly, and changed the economics after halting new launches. WALLET followed the public path; no Noxa-specific allocation, override, or pre-launch top-level contact was located.
The cohort adds a directly supportive comparative fact. Only 11 of 151 surrounding launches had no creator initial buy, and WALLET alone linked to a page on Robinhood's strict official domain. The official-domain metadata was user-entered, so it cannot establish authorship. Combined with locked base liquidity, no creator token allocation, fee burning, and aa33's separate provenance, it makes the clean community-launch case harder to dismiss.
The verified ERC-8056 source adds product specificity: aa33 described the Base contract as a Mercury integration test. Robinhood Wallet now executes perpetuals through Lighter on Robinhood Chain, confirming that the wallet has become an onchain trading surface. Mercury remains unattributed, and neither fact proves WALLET token utility.
The official-or-fake binary still fails. The evidence supports a professionally built community asset launched by an exceptional deployer. aa33's demo-wallet funding, named ERC-8056 integration test, restricted-mainnet activity, and F718 connection remain the central Robinhood-proximity signals that could move this from a strong circumstantial thesis to attribution.
Four tests could take the case beyond circumstantial evidence.
Source ledger
Method: MegaETH and Robinhood Chain JSON-RPC supplied runtime bytecode. MegaETH Etherscan, Robinhood Chain Blockscout, Base Blockscout, and Ethereum Blockscout supplied deployment records, verified source, compiler metadata, decoded calldata, transaction receipts, token transfers, current state, NFT ownership, balances, and indexed transaction history. Relay request status was matched to each of the 14 origin hashes; settlement fills and refunds were then checked against destination-chain transactions. The launch cohort contains the 75 successful launchToken calls immediately before WALLET, WALLET itself, and the next 75 calls. Fee percentages and cohort values were calculated from raw integers. Treasury reconciliation excludes two reverted transactions. RAILGUN shielding establishes a public trace boundary; no private-pool beneficiary is inferred without a viewing key. Public-source searches did not identify Mercury or verify the dYdX/Arcus and 50/50 revenue claims. Media reporting is used for public timing and reported scale, not contract mechanics.