---
name: operator-graph
description: Operator graph on Robinhood Chain: deployer, funder and wallet links between projects. (frontmatter rebuilt during handover copy)
sources: [cowork]
---
# Operator graph, Robinhood Chain

- [stated] Ben approved building this on 31 Aug 2026, agreeing operators repeat and prices do not.

## Why this table exists
Operator behaviour is persistent across launches. Price is not. A funder that has bankrolled three deployers, one of which extracted 8.765 ETH, is a fact available at hour zero of the next launch. Nothing else in the stack carries forward between screens.

## Method
Dune, standing query id 8562361. Key and gotchas in `research-sources` section 5b.
- Launch set: `robinhood.transactions` where `"to" = <factory>` and success.
- Funder: earliest inbound value transfer to each deployer.
- Extraction: ETH to RelayDepository `0x4cD00E387622C35bDDB9b4c962C136462338BC31`, direct plus one hop, because operators route through single-use EOAs.
- Fees: `robinhood.traces` from the launchpad fee router.
**Blockscout undercounts internal transfers. Dune read 9.215 ETH of fees where Blockscout's internal-tx list showed 8.9243. Use Dune for any cumulative figure.**

## Known infrastructure addresses
- v4.fun factory / deployer `0xd541cd01c8C6C68e7415A7324e060f4c250363A3`
- v4.fun creator fee router `0xCa3ea6d93aE7Ec71edC74AbaaCCd8bA736373045`, method `claimCreator`
- RelayDepository `0x4cD00E387622C35bDDB9b4c962C136462338BC31`, the standard exit route off this chain
- Uniswap v4 PoolManager `0x8366a39CC670B4001A1121B8F6A443A643e40951`

## Funders, ranked by deployers funded (60 day window, 31 Aug 2026)

| Funder | Deployers funded | Active since |
|---|---|---|
| `0xa67d7eb4dc68fa6ce8e34ef8cadaf075b9893fbb` | **3** | 14 Jul 2026 |
| `0xf70da97812cb96acdf810712aa562db8dfa3dbef` | 2 | 4 Jul 2026 |
| `0xabb2acd3be814a80e502575d6c1dc5f789e9cd10` | 2 | 21 Jul 2026 |

**7 of 15 v4.fun deployers in 60 days trace back to 3 funders.**

## Deployers

| Deployer | Token | Fees in (ETH) | Bridged to Relay (ETH) | First launch |
|---|---|---|---|---|
| `0x7ff83a35882436f9556c69a6c5c249d58fac3bab` | SCL / Scalar | 9.215 | 8.444 | 29 Aug 20:47 |
| `0xba545fbc19b5d141a0166836b47d8432d844a7ae` | unresolved | 0 (different router) | **8.765** | 29 Aug 19:17 |
| `0xb451b975731bc04302b06e68ddace5e04b947b89` | unresolved | 0 | 4.222 | 29 Aug 18:16 |
| `0x45b54ba34b84b21f61ddd3f0d80d3ccb3d4be895` | unresolved | 0 | 0 | 29 Aug 18:19 |
| `0xb919e476dbe6f3a85985b64f26e0e895fa96a2e3` | unresolved | 0 | 0.548 | 31 Aug 05:59 |
| `0x040db6261db1961a47014c9e6dd40ca48863a70e` | unresolved | 0 | 0.103 | 30 Aug 06:31 |
| `0x89695601d06a23d31561a929f14b0a520c233716` | unresolved | 0 | 0.058 | 30 Aug 15:45 |
| `0x0bb3af56aca11237a5e7e8cf1c2ed395b0095721` | unresolved | 0 | 0.074 | 29 Aug 21:42 |
| `0x22190fef31d75d8d5a8bc8c6a0aea965c50b17f0` | unresolved | 0 | 0 | 31 Aug 08:29 |
| `0x3993081ca76a8341d833c34871ca9c5324e42b94` | unresolved | 0 | 0 | 29 Aug 18:56 |
| `0xfbadfaba4c977e2dfae45fa8de44e7555973397e` | unresolved | 0 | 0 | 30 Aug 06:58 |
| `0xf44510ca14f2f3876187784859eb2f91d3337f86` | unresolved | 0 | 0 | 30 Aug 17:05 |
| `0x993d9c7087c99b16d4a118feff583610da1e6cea` | unresolved | 0 | 0 | 30 Aug 05:12 |
| `0xd4adddfba6fce9c993e432f39403e8ddcd4b1796` | unresolved | 0 | 0 | 29 Aug 14:51 |
| `0x3a8805b001147d364308c866e4b6734efac28eec` | unresolved | 0 | 0 | 29 Aug 18:24 |

## THE DEGREE FILTER, learned the hard way on the first run, 31 Aug 2026
The first funder query looked like a major finding: `0xa67d7eb4dc68fa6ce8e34ef8cadaf075b9893fbb` first-funded the SCL deployer AND `0xba545fbc19b5d141a0166836b47d8432d844a7ae` (which bridged 8.765 ETH to Relay) AND `0x45b54ba34b84b21f61ddd3f0d80d3ccb3d4be895`, active since 14 Jul 2026.

**That reading was FALSE and was killed by the chain wide degree test.** Chain wide, that address has sent ETH to **91,649 addresses** and was the FIRST funder of **46,398** of them, totalling 8,834.9 ETH. It is infrastructure (a bridge payout, exchange hot wallet or onboarding wallet), not an operator. It is evidence about nothing.

**STANDING RULE. Never attribute a funder to a project until its chain wide out-degree is measured.**
- Out-degree above roughly 100 first-funded addresses: infrastructure. Discard, do not report.
- Out-degree in single or low double digits: a real operator, worth the table.
- This is the `lowcap-screen` second-token rule applied to funders. Skipping it produces a confident and completely wrong story, which is exactly what happened on the first run.

The settling query:
```
with sends as (select "to" as r, min(block_time) t from robinhood.transactions
               where "from" = <FUNDER> and value > 0 and success group by 1),
their_first as (select t."to" r, min_by(t."from", t.block_time) ff from robinhood.transactions t
                where t.value > 0 and t.success and t."to" in (select r from sends) group by 1)
select count(*) sent_to, count_if(f.ff = <FUNDER>) first_funded from sends s join their_first f on f.r = s.r
```

## OPEN, CLOSED 1 Sep 2026
`0xf70da97812cb96acdf810712aa562db8dfa3dbef` and `0xabb2acd3be814a80e502575d6c1dc5f789e9cd10` were unresolved. **Both were degree-tested on 1 Sep and both FAIL. Neither is an operator. Discard both.**

| Funder | Deployers funded (3d window) | Addresses sent to | First-funded | Verdict |
|---|---|---|---|---|
| `0xa67d7eb4dc68fa6ce8e34ef8cadaf075b9893fbb` | 3 | 91,649 | 46,398 | infrastructure, killed 31 Aug |
| `0xf70da97812cb96acdf810712aa562db8dfa3dbef` | 3 | **145,708** | **105,696** | infrastructure, killed 1 Sep |
| `0xabb2acd3be814a80e502575d6c1dc5f789e9cd10` | 3 | **84,990** | **40,457** | infrastructure, killed 1 Sep |

**ALL THREE multi-deployer funders on this chain are now infrastructure. The funder axis has produced zero operators in two runs.** The "7 of 15 deployers trace back to 3 funders" headline above is therefore a statement about three hot wallets, not about three operators, and should not be quoted as a finding.

**SECOND-ORDER USE FOUND 1 Sep.** `0xf70da97812cb96acdf810712aa562db8dfa3dbef` also appears in the ARC screen's top-15 buyer list, called a "spray bot, early into 270 tokens". At 145,708 out-degree it is not a trader at all. **The degree filter applies to BUYER wallets as well as funders. Apply it before calling any wallet a spray bot, a smart-money wallet or an elite entry.** Correction written back into [[screen-log-outcomes]].

## Run of 1 Sep 2026, v4.fun factory, 3 day window
19 first-funding rows returned. **No repeat operator: not one new deployer's funder survives the degree filter, and no new deployer was already in the table.**

**IMPORTANT COLUMN CAVEAT.** The 1 Sep query measures relay ETH **direct only**, no hop. The original table column is direct plus one hop and is NOT superseded. Same wallets read differently: `0x7ff83a35882436f9556c69a6c5c249d58fac3bab` 8.444 with the hop, **0 direct**; `0xba545fbc19b5d141a0166836b47d8432d844a7ae` 8.765 with the hop, 8.114 direct; `0xb451b975731bc04302b06e68ddace5e04b947b89` 4.222 with the hop, 1.600 direct. **A direct-only query reads the SCL operator as clean. Always run the one-hop version before clearing a deployer.**

### New deployers, first seen 1 Sep 2026
| Deployer | Funder | Relay out, direct only (ETH) | Funder status |
|---|---|---|---|
| `0xe75c7c4e65e63ec79f9c91338b2d9704b8f61771` | `0xf70da97812cb96acdf810712aa562db8dfa3dbef` | 1.2201 | infrastructure, no attribution |
| `0xef2fb4861312774ecd423e6d9b3c0dbd7bf9377e` | `0xabb2acd3be814a80e502575d6c1dc5f789e9cd10` | 0.2270 | infrastructure, no attribution |
| `0xfdd07af7f92ad22175ccc6b83f41e08f2112b302` | `0xa5df4d576a72ddbbb4296614aaf58ed6696ebe61` | 0 | single, untested |
| `0xd478e88fae5ab604b31c77ccf3e5d9d4edfe1679` | `0xa5a5491bca93dd4c076e4906e79e7673f4a5a142` | 0 | single, untested |
| `0x9c9a205e852a6759fad68476aed8ef1537c5b1cd` | `0x2acc803c312980dae95c0bcce08baaf9365053c6` | 0 | single, untested |

`0xe75c7c4e65e63ec79f9c91338b2d9704b8f61771` bridged **1.2201 ETH direct to RelayDepository** and is the only new deployer with a material extraction. Token not yet resolved. Resolve it before it appears in a screen.

`0xa5a5491bca93dd4c076e4906e79e7673f4a5a142` also appears in the ARC screen's spray-bot list (100 tokens). Untested for degree, so **do not attribute**, same rule as above.

### Funders newly seen, 1 each, all untested
`0x88766f7d78004dd5b2556b6d88bab685859dde5b` (funded `0xb451b975731bc04302b06e68ddace5e04b947b89`), `0xa5df4d576a72ddbbb4296614aaf58ed6696ebe61`, `0xa5a5491bca93dd4c076e4906e79e7673f4a5a142`, `0x2acc803c312980dae95c0bcce08baaf9365053c6`, `0x74664d5cfda0fedb5a0c67cf382202211a112e8c`, `0x712850b2400019f583cbbd4d3aa034b46bd79ec8`, `0x17e6eeb4efcc48a70f36201fc3cd3ee65ed4c77f`, `0xb484526906b9512fc055ffde71c64c2567b9edbb`, `0xee8fac3c6fb4f52f626cc4d8343261fbf03763f2`, `0x56c262027e0de4aea31d2489529cb25d23e58a8b`

`0x3a8805b001147d364308c866e4b6734efac28eec` (29 Aug 18:24) fell out of the 3 day window and did not reappear.

## Pons launchpad, a SECOND fee architecture, found 31 Aug 2026 on DICE

The v4.fun table above is not the whole chain. Tokens launched through **Pons** use a different set of contracts and a fee stream **paid in the project's own token, not ETH**, so a Relay-bridging query will never see the extraction.

- `PonsLaunchFactory` deployer of the token contract: `0xA5aAb3F0c6EeadF30Ef1D3Eb997108E976351feB`
- `PonsLaunchLocker` `0x736d76699c26d0d966744cae304c000d471f7f35`, creator of record `0xda4bCee76B29EFEc9697Fcf663601c2042043968`
- Token template: `PonsLauncherToken`, fixed supply minted to the factory, launch-block and maxWallet protections that expire at `restrictionEndBlock`, then a plain ERC-20 with no owner and no mint

**`0xda4bCee76B29EFEc9697Fcf663601c2042043968` fails the degree filter: 54,708 transactions and 1,033,659 token transfers. It is Pons infrastructure, not an operator. Do not attribute it to any project.**

### DICE operator row
| Field | Value |
|---|---|
| Token | DICE / Dice Protocol `0x3F9f0b6073Ee8c495Aed96869AF31850fED40FeB` |
| Launchpad | Pons (V3 launch), graduated |
| Fee-collection EOA | `0x4acd2c88a239a924e47fc4995114ca1bb0ca3cad`, **zero normal transactions**, only receives |
| Fees in | 48,930,665 DICE from PonsLaunchLocker, 29 Jul to 10 Aug 2026, in a drip of 90+ transfers |
| Burned | 26,010,573 DICE (2.601%) to 0xdEaD in five transfers 1 to 5 Aug. **Cost $0**, it is free fee inventory |
| Forwarded out | 22,920,092 DICE (2.292%) to `0xd2057832bc4abfb4aa0d76d5ba160a079554cc10` 8,275,844 · `0x46d7aba22ace55596d7f42b0db2b068a5ca5e5b6` 7,936,896 · `0xa18a36a794048f4cd18e188b9896834a2d8e7c04` 4,360,734 · `0x918eaf0b2589710b0d85ef48c12a343e68263841` 2,246,618 · Provider `0x8741b8a825644D9Ef18Faf2DAB5e9b47B900F2b6` 100,000 |
| Bridged to Relay | not measured, the fee stream is in DICE not ETH |
| Last activity | 10 Aug 2026 |

**STANDING RULE ADDED: on a Pons launch, measure the creator fee stream in the PROJECT'S TOKEN out of PonsLaunchLocker. The v4.fun ETH-to-Relay method returns zero and reads as a clean operator when it is not.** A burn financed from this stream must be priced at $0 and must never be reported as a buyback.

### BACKED operator row, 31 Aug 2026
| Field | Value |
|---|---|
| Token | BACKED `0x7168563B0E70124f0C7c0cF2F13a8D1861BAf4A5`, Robinhood Chain |
| Deployer / vault owner | `0x9ca5B125C8369df97642C012bE2B356c3165DCC7`, plain EOA, no timelock |
| Launchpad | none. Self-deployed with a mined CREATE2 salt for the v4 hook permission bits |
| Deployer token balance | 13,000,000 BACKED = 1.377% of supply |
| Deployer ETH | 1.2405 |
| Powers held | StockVault owner: setKeeper, addStock, removeStock, pruneRedeemable (zero-balance only), setRedemptionFeeBps (capped 10%), setMaxSpendPerBuy. NO path to the accumulated stocks |
| Residual drain vector | addStock with an arbitrary token and hook, then keeper buyStocks spends balance/n through it. Exposure is the un-swept ETH between sweeps, $325 at screen |
| Fees extracted | ZERO measurable. 100% of the 3% tax is in the vault, reconciled to within 3.3% of lifetime volume |
| Bridged to Relay | none found |
| LP | locked at UNCX `0x128A800cBc615cc110Bff16E475865c67631603A` lock id 7 to 30 Jun 2027 15:27 UTC, receipt NFT held by this EOA |
| Second token | DENAR `0x3786728a2c49C4617Bf4FE5BD82b90B6B0dF5508`, PonsV2LauncherToken, launched 30 Aug 2026 21:37 UTC, $1,938,184 at 14h old. Same team by their own announcement. **Apply the Pons creator-fee rule above: the extraction stream will be in DENAR, not ETH, so a Relay query returns clean.** Not yet measured |

**This is the cleanest operator row in the table so far: no launchpad fee stream, no bridged ETH, no dev sell path, LP locked with a date, and a treasury that reconciles to the tax within 3.3%.** Use it as the positive control when a screen's operator behaviour looks ambiguous.

### DEBT operator row, 31 Aug 2026
| Field | Value |
|---|---|
| Token | DEBT `0x5407B93692B5FCD2BBf5F8CfB328fAE68006F786`, Robinhood Chain, PonsV2LauncherToken |
| Launchpad | Pons V2, graduated to a Uniswap v4 ETH pool the same day |
| Fee-recipient EOA ("the Obligor") | `0xc7926ca2Ba6aF6FcdF3a627687a9363885c9048F`, plain EOA, claims from V2FeeEscrow `0xd3AFEB2a57f70eF218Aa82451c51B2fb0416Ac9e` |
| First funder | `0xca7ded7e4F4Ba8aB3B10009236Ae6d1b95094589`, 0.0507 ETH at 06:27 UTC. **DEGREE TEST NOT RUN, do not attribute** |
| Fee stream currency | **ETH, not the project token.** Pons V2 escrow pays ETH, unlike the Pons V1 DICE case where the stream was in DICE |
| Fees in | 0.5377 ETH lifetime ($1,330) in the first 8.4 hours |
| Extracted | ZERO so far. 0.3226 ETH paid out as pro-rata coupons to 683 holders, 0.1613 ETH spent on open-market buybacks and burned |
| Residual held | 0.5389 ETH in the EOA against a stated 0.0564 ETH reserve. Un-swept, and sweepable at will |
| Bridged to Relay | none found |
| Drain vector | **the whole mechanism.** The waterfall is a process on debt-engine.fly.dev holding this key. No contract enforces the 60/30/10 split, the payer address, or continuation |
| Prior issuance | a retired Solana launch, evidenced by app.js purging a cached base58 mint and by the whitepaper still saying "SPL token" and "SPL burn" |

**NEW OPERATOR PATTERN, name it "the off-chain covenant".** A token whose value accrual is real, verifiable and running, but implemented as a bot rather than a contract. It reads clean on every extraction query (no bridged ETH, no dev bag, no token-denominated fee stream) precisely because the operator is currently choosing to pay. The measurement to take is not "has he extracted" but "what is un-swept right now and what stops him". Contrast BACKED, where the residual sat in a contract the deployer had no path into. Here it sits in the deployer's own EOA.
---

## Run of 2 Sep 2026, v4.fun factory, 3 day window. **THE GRAPH FINALLY PAID.**

24 first-funding rows returned, 9 deployers not previously in this table. **Three funders SURVIVE the degree filter. These are the first genuine operators the funder axis has produced in three runs, and one of them is already in this table as a deployer.**

### DEGREE TEST RESULTS, 2 Sep 2026
| Funder | Sent to | First-funded | Verdict |
|---|---|---|---|
| `0x56c262027e0de4aea31d2489529cb25d23e58a8b` | 95,795 | 48,372 | infrastructure, killed 2 Sep |
| `0xf42b73d2e4d3912eade2a5f7ab39b54064c2acec` | 2,704 | 879 | infrastructure, killed 2 Sep |
| `0x2f075cbdce9c2e776bb720f7d5300e9c26915983` | 9 | **4** | **OPERATOR** |
| `0xba545fbc19b5d141a0166836b47d8432d844a7ae` | 11 | **3** | **OPERATOR, and already a deployer in this table** |
| `0x18c2ff18dcec0d3ca5d338ed68bab88a3dbebc5c` | 15 | **3** | **OPERATOR** |
| `0x1bfa5b1984efa66df57fe39eccefde1dcd520ceb` | 1 | 1 | single, no cluster |
| `0xe71760d86e7b72136dc4ab5e3482a54c289953c3` | 1 | 1 | single, no cluster |

`0x56c262027e0de4aea31d2489529cb25d23e58a8b` was listed on 1 Sep as an untested single funder. It is a hot wallet, 48,372 first-funded. Struck.

### CLUSTER A, THE REPEAT OPERATOR. Funder `0xba545fbc19b5d141a0166836b47d8432d844a7ae`
**This wallet is row 2 of the deployer table above: launched on v4.fun 29 Aug 19:17 and bridged 8.765 ETH to RelayDepository (8.114 direct). It is now confirmed as the funder of three other wallets, and it passes the degree filter at 11 sent to / 3 first-funded. An extractor bankrolling new launches is exactly the hour-zero signal this file was built for.**

| Child | First funded | v4.fun launches | Relay direct (ETH) | Relay one-hop (ETH) |
|---|---|---|---|---|
| `0x31511477f9d0a45a65f7e0a211749573fb5e7b7f` | 2026-08-08 20:17:05 | **11** | 0 | 0 |
| `0x0ccacfa562f72b667a014183abf0115ed88a367b` | 2026-08-10 15:55:29 | 0 | 0.150 | 0 |
| `0x2537c9e19ddda62aa0a034782b9eecc8de6c49ea` | 2026-08-23 20:42:03 | 0 | 0.501 | 0 |

- `0x31511477f9d0a45a65f7e0a211749573fb5e7b7f` has run **11 v4.fun launches** and has extracted nothing yet. It is the production wallet of this cluster
- the other two are pure exit wallets, 0.651 ETH out between them, no launches
- cluster total bridged to Relay: **8.765 ETH** by the funder itself plus 0.651 by the children
- **ACTION: treat any token deployed by `0x31511477f9d0a45a65f7e0a211749573fb5e7b7f` or `0xba545fbc19b5d141a0166836b47d8432d844a7ae` as pre-flagged at hour zero. Eleven launches from one wallet funded by an 8.765 ETH extractor is a serial launcher, not a founder**

### CLUSTER B. Funder `0x2f075cbdce9c2e776bb720f7d5300e9c26915983`, 4 first-funded
| Child | First funded | v4.fun launches | Relay direct (ETH) | Relay one-hop (ETH) |
|---|---|---|---|---|
| `0xac351e530d0bd83411b0159a2c68c9cbfe39fe12` | 2026-08-13 17:08:15 | 0 | 0.005 | **4.5417** |
| `0x9364a75257bcac28dc1ced884b7cf8dc7db2221b` | 2026-08-14 06:20:32 | 0 | 0.0149 | 0 |
| `0x76d64eb70f88ff4cfbdaaf012a5e559e56d2ca11` | 2026-08-30 17:15:24 | 0 | 0 | 0 |
| `0xbc2aaca7cedfc9c07d66d19444f4f87282a8dd22` | 2026-08-30 19:44:55 | **1, new this run** | 0 | 0.010 |

- **4.5417 ETH bridged out through one hop by `0xac351e530d0bd83411b0159a2c68c9cbfe39fe12`.** The direct-only column reads 0.005 and would have cleared this wallet. Another instance of the one-hop caveat below
- cluster total bridged: **4.567 ETH**. New launch today from `0xbc2aaca7cedfc9c07d66d19444f4f87282a8dd22`

### CLUSTER C. Funder `0x18c2ff18dcec0d3ca5d338ed68bab88a3dbebc5c`, 3 first-funded
| Child | First funded | v4.fun launches | Relay direct (ETH) | Relay one-hop (ETH) |
|---|---|---|---|---|
| `0x209ca052e7087a2f6f3e5ed4558f5908ad8928da` | 2026-07-10 13:06:25 | 0 | 0 | **1.830** |
| `0x0dfd581cc618fa721083859364c707bb9fe68127` | 2026-07-15 13:36:17 | 0 | 0 | **1.513** |
| `0x758d3b0c79a032cfd83c4315d071d333bebf203e` | 2026-08-06 09:08:50 | **2, new this run** | 0 | 0 |

- cluster total bridged: **3.343 ETH, all of it one-hop, zero direct.** New launches today from `0x758d3b0c79a032cfd83c4315d071d333bebf203e`
- oldest cluster in the set, active since 10 Jul 2026

### New deployers this run, not previously in the table
| Deployer | Funder | Relay direct (ETH) | Funder status |
|---|---|---|---|
| `0xcdbb76d1d731b268c090f18603c888b3f75c8a58` | `0x56c262027e0de4aea31d2489529cb25d23e58a8b` | **1.9099** | infrastructure, no attribution |
| `0x31511477f9d0a45a65f7e0a211749573fb5e7b7f` | `0xba545fbc19b5d141a0166836b47d8432d844a7ae` | 0 | **OPERATOR, cluster A** |
| `0xbc2aaca7cedfc9c07d66d19444f4f87282a8dd22` | `0x2f075cbdce9c2e776bb720f7d5300e9c26915983` | 0 | **OPERATOR, cluster B** |
| `0x758d3b0c79a032cfd83c4315d071d333bebf203e` | `0x18c2ff18dcec0d3ca5d338ed68bab88a3dbebc5c` | 0 | **OPERATOR, cluster C** |
| `0x50492f0ecdeb4bf76f9387b3b26c3af1656cc2c6` | `0x56c262027e0de4aea31d2489529cb25d23e58a8b` | 0 | infrastructure, no attribution |
| `0xf177b7f781292faa6d224c4a4fdefb1eae80ad2e` | `0xe71760d86e7b72136dc4ab5e3482a54c289953c3` | 0 | single, no cluster |
| `0x9d3308f62ab6eda8e868a2300568e8b51edb1131` | `0x1bfa5b1984efa66df57fe39eccefde1dcd520ceb` | 0 | single, no cluster |
| `0x7641989cec1694f4d4547370b4312ff98e984d2c` | `0xf42b73d2e4d3912eade2a5f7ab39b54064c2acec` | 0 | infrastructure, no attribution |
| `0x943f9ff0c22149f6e4732b7a5a7205b66f37e52b` | `0xf70da97812cb96acdf810712aa562db8dfa3dbef` | 0 | infrastructure, no attribution |

### Existing deployers, extraction re-read 2 Sep (direct only)
- `0xb919e476dbe6f3a85985b64f26e0e895fa96a2e3` now **1.549 ETH direct**, against 0.548 recorded 31 Aug on the direct-plus-hop method. **It has kept extracting.** Re-measure with the hop before quoting a cumulative figure
- `0xba545fbc19b5d141a0166836b47d8432d844a7ae` 8.114 direct, unchanged
- `0xb451b975731bc04302b06e68ddace5e04b947b89` 1.600 direct, unchanged. Funder `0x88766f7d78004dd5b2556b6d88bab685859dde5b`, still untested
- `0xef2fb4861312774ecd423e6d9b3c0dbd7bf9377e` 0.227, `0x040db6261db1961a47014c9e6dd40ca48863a70e` 0.0466, `0x0bb3af56aca11237a5e7e8cf1c2ed395b0095721` 0.0160, `0xe75c7c4e65e63ec79f9c91338b2d9704b8f61771` 1.2201, all unchanged

### WHAT CHANGED IN THE METHOD, 2 Sep 2026
1. **The degree filter is not only a killer, it is a finder.** Two runs of "everything is infrastructure" made the funder axis look worthless. It is not: the filter needed a wider window and the one-hop relay column to separate the three real clusters from the three hot wallets. **Do not retire the funder axis.**
2. **A funder that is ITSELF a deployer in this table is the strongest signal available and costs one join.** Add it as a standing check: after every run, intersect the new funder set against the existing deployer column. That single line is what found cluster A.
3. **The one-hop column decided two of three clusters.** Cluster B's 4.5417 ETH and cluster C's entire 3.343 ETH are invisible to a direct-only query. The 1 Sep caveat is now confirmed twice over: **never clear a wallet on a direct-only read.**
4. Standing query 8562361 runs about 100 to 110 seconds at medium performance. Three PATCH-execute-poll cycles fit comfortably in one session. Poll with nohup, a foreground poll loop hits the 2 minute bash timeout.

**DEBT ROW UPDATE, 2 Sep 2026: the off-chain covenant was exercised within 14 hours.** 0.4 ETH to `0xaf924bbC01686d0E89e00CB69C079dAA8A3c595c` (EOA, 12.43 ETH, active sniper via `0x4A86009A36FceC5Aa341ffCEB3205a911FcF6f60`) and 0.404 ETH to `0x9e48C2143B21c80eF905a7F46c7cF3b60B48cc0f`. Same key then launched rob `0x1166C139C86BcEc77B4E37941dCd1Df04C9401AC`, ROBBIE `0x565801ff331505c67922e32b8F203fb558E329f1`, ROBBIE `0x9317977bC68B01B52df5A90ae3A397E8840A6419`, rob `0x96fa1698c4D17b24aF40835fec5fE60f7AEEe9f9` on PonsV2LaunchFactory `0x7eD598BcEf8bd9Edd8C97A195C6d13f40801EC7e`. **Add `0xaf924bbC01686d0E89e00CB69C079dAA8A3c595c` and `0xc7926ca2Ba6aF6FcdF3a627687a9363885c9048F` to the deployer watch; any token whose creator fee recipient resolves to either is this operator.** Degree test on `0xaf924bbC` still to run.

---

## Run of 3 Sep 2026, v4.fun factory, 3 day window. **THE PAD WENT TO ZERO.**

20 first-funding rows. **Every one of the 20 deployers is already in this table. Zero new deployers, zero new funders, zero new repeat operators.**

### THE FINDING OF THIS RUN: v4.fun has had NO launches for 29 hours
Second query, launch counts per creator, plus a freshness check:
- **`v4fun_24h` = 0.** Not one successful transaction to factory `0xd541cd01c8C6C68e7415A7324e060f4c250363A3` in the last 24 hours
- `v4fun_48h` = 31, so every launch in the window happened in the first half of it
- **last v4.fun launch: 2026-09-01 20:20:41 UTC by `0x040db6261db1961a47014c9e6dd40ca48863a70e`**
- **NOT a Dune ingestion lag, checked explicitly**: `max(block_time)` on `robinhood.transactions` = 2026-09-02 23:17:12 UTC (about 2h behind wall clock) with **11,409,541 transactions in the last 24h**. The chain is busy; the pad is not
- **PonsV2LaunchFactory `0x7eD598BcEf8bd9Edd8C97A195C6d13f40801EC7e` took 10,603 transactions in the same 24 hours.** DefiLlama agrees: Pons V2 fees24h $5,702,128, chain rank #2; o1 Launchpad $411,118, rank #6; Pons V1 $335,884; NOXA Fun $153,498
- **THE METHOD CONSEQUENCE, and it is the important one. This file's launch set is `"to" = <v4.fun factory>`. That axis is now dead and the graph will keep returning "nothing new" forever while the actual launch flow runs through Pons V2. Add PonsV2LaunchFactory `0x7eD598BcEf8bd9Edd8C97A195C6d13f40801EC7e` as a second launch set before the next run, and treat the v4.fun set as legacy.** The Pons fee-architecture section below already documents that the extraction stream there is token-denominated (V1) or ETH via V2FeeEscrow `0xd3AFEB2a57f70eF218Aa82451c51B2fb0416Ac9e`, so the relay-ETH column needs the V2 escrow path, not the v4.fun path

### Launch counts in the window, all 20 creators already in the table
| Deployer | Launches 3d | Last launch (UTC) | Status |
|---|---|---|---|
| `0x31511477f9d0a45a65f7e0a211749573fb5e7b7f` | **11** | 2026-09-01 19:02:55 | cluster A production wallet |
| `0x943f9ff0c22149f6e4732b7a5a7205b66f37e52b` | **7** | 2026-09-01 19:34:50 | funder is infrastructure, no attribution |
| `0xef2fb4861312774ecd423e6d9b3c0dbd7bf9377e` | 4 | 2026-08-31 17:35:31 | funder is infrastructure |
| `0xba545fbc19b5d141a0166836b47d8432d844a7ae` | 2 | 2026-09-01 14:45:29 | **cluster A funder AND deployer** |
| `0x758d3b0c79a032cfd83c4315d071d333bebf203e` | 2 | 2026-09-01 17:05:57 | cluster C child |
| `0xb451b975731bc04302b06e68ddace5e04b947b89` | 2 | 2026-09-01 19:04:52 | funder `0x88766f7d78004dd5b2556b6d88bab685859dde5b`, still untested |
| `0xbc2aaca7cedfc9c07d66d19444f4f87282a8dd22` | 1 | 2026-09-01 15:26:42 | cluster B child |
| 13 others | 1 each | 2026-08-31 to 2026-09-01 | all previously recorded |

**The three clusters found on 2 Sep are all still the whole story: A (`0xba545fbc19b5d141a0166836b47d8432d844a7ae` + `0x31511477f9d0a45a65f7e0a211749573fb5e7b7f`), B (`0x2f075cbdce9c2e776bb720f7d5300e9c26915983` + `0xbc2aaca7cedfc9c07d66d19444f4f87282a8dd22`), C (`0x18c2ff18dcec0d3ca5d338ed68bab88a3dbebc5c` + `0x758d3b0c79a032cfd83c4315d071d333bebf203e`). Nothing new joined them and none of them launched in the last 24 hours either.**

### Extraction re-read 3 Sep, direct only
- **`0x943f9ff0c22149f6e4732b7a5a7205b66f37e52b` now 0.59 ETH direct, up from 0 on 2 Sep.** Seven launches in the window and it has started extracting. Its funder `0xf70da97812cb96acdf810712aa562db8dfa3dbef` is infrastructure (145,708 out-degree), so there is no attribution, but **the wallet itself is now the second-most active launcher on the pad AND a live extractor. Degree-test it as a funder and run the one-hop relay column on it next run**
- unchanged: `0xba545fbc19b5d141a0166836b47d8432d844a7ae` 8.1139 · `0xcdbb76d1d731b268c090f18603c888b3f75c8a58` 1.9099 · `0xb451b975731bc04302b06e68ddace5e04b947b89` 1.600 · `0xb919e476dbe6f3a85985b64f26e0e895fa96a2e3` 1.549 · `0xe75c7c4e65e63ec79f9c91338b2d9704b8f61771` 1.2201 · `0xef2fb4861312774ecd423e6d9b3c0dbd7bf9377e` 0.227 · `0x040db6261db1961a47014c9e6dd40ca48863a70e` 0.0466 · `0x0bb3af56aca11237a5e7e8cf1c2ed395b0095721` 0.0160
- `0x31511477f9d0a45a65f7e0a211749573fb5e7b7f` still 0 direct after 11 launches. Eleven launches and nothing extracted is itself a state worth watching: either the launches failed to raise, or the extraction is queued
- **still owed from 2 Sep: the degree test on `0xaf924bbC01686d0E89e00CB69C079dAA8A3c595c` (the DEBT off-chain-covenant payee). Not run for a second time**

### WHAT CHANGED IN THE METHOD, 3 Sep 2026
1. **A launch-count query alongside the funding query costs one extra execution and is what turned "20 rows, nothing new" into "the pad is dead".** The funding query alone returns rows for creators whose launches happened days ago and reads as ongoing activity. **Always pull `count_if(block_time > now() - interval '1' day)` per creator, not just the 3-day set.**
2. **Always check `max(block_time)` on the source table before reporting a zero.** A zero from a lagging table and a zero from a dead pad look identical. Confirmed live here at 11.4M transactions in 24h.
3. Query 8562361 ran in **80 to 110 seconds** at medium performance, three PATCH-execute-poll cycles in the run. Nohup poll with a 10s interval; a foreground loop still hits the bash timeout.

---

## CONTINUED IN [[operator-graph-2]] FROM 4 SEP 2026
This file is near the 32KB cap. Runs from 4 Sep 2026 onward are in [[operator-graph-2]]. **Read both.**
Headline of the 4 Sep run, recorded here because it overturns the 3 Sep conclusion above: **v4.fun is NOT dead. Two launches on 3 Sep after a 41 hour gap, by `0x39aec46f5ec95169dae94c0f2d3c9f6c97893488` and `0x1fb3bfabf44e511546499b7a581a4b4684adbd76`. Do not retire the v4.fun launch set; add the Pons V2 set alongside it.** Also closed on 4 Sep: `0xaf924bbc01686d0e89e00cb69c079daa8a3c595c` passes the degree filter at 21 sent to / 9 first-funded and is a **confirmed operator**.