# A1.14 Where v3 coins come from (Analyst 1, 6 Oct 2026, 15:55 Bangkok)

## Read this first
- **The coins that reach the $1M gate come from two copy-paste "launch kits", not from a launchpad.** Each kit is a small contract with identical bytecode. The dev deploys their own copy of the kit, then calls it seconds later to create the token, the pool and the liquidity.
  - **Kit A** (bytecode 5,563 bytes, hash 0e81be52d5cf): 20 copies, 107 launches, 24 Sep to 1 Oct. **44 reached the gate (41%).** It has stopped.
  - **Kit B** (bytecode 5,050 bytes, hash 89885b75fd29): a fresh copy for every launch, 140 launches since 1 Oct (14 to 50 a day). **28 reached the gate (20%).** It is active today.
  - Everything else: 6,861 launches, 42 reached the gate (0.6%). The Uniswap position manager `0x73991a25c818bf1f1128deaab1492d45638de0d3` (the plain "add liquidity" route) made 4,785 launches but only 18 gate coins (0.4%), and only 1 of them in the last 7 days.
- **Is the flow growing or drying up? Flat in count, and it now hangs on one kit.**
  - Wide-rule coins per day: 4.9 (21 to 29 Sep) vs 3.7 (30 Sep to 6 Oct; 6 Oct is a part day). All v3 WETH launches: about 300 a day, flat.
  - **In the last 7 days, 15 of the 26 wide-rule coins came from Kit B (58%), 6 from Kit A, 1 from the position manager and 4 from others.** Half A was mostly the position manager (17 of 36).
  - Money per launch source on the wide rule (all 16 days): Kit A +$736 a day, Kit B +$675, position manager +$684, others +$384. **Kit B coins average 2.44x**, the best of any source.
  - **Risk:** if Kit B stops, more than half of today's flow goes with it.
- **How Kit B launches are funded (25 of 25 checked):** one contract, `0x84157d3d24630c297849c84d49ae7f1805c1b60f` (9,387 bytes, has disperse and swap functions), sends ETH to a brand-new deployer wallet 1 to 2 s before the launch.
  - The deployer gets 4 to 28 ETH (median about 12). In the same transaction, 2 to 13 other fresh wallets get exactly 0.01 ETH each, which is gas money.
  - A different caller sends each funding transaction. I can't prove from this whether that is one operator or a shared tool.
  - **Those same 0.01 ETH wallets buy 26% to 38% of the pre-gate ETH** in the 4 Kit B coins I could check. That is the dev's own bundle buy, and it counts toward "sellable $1M".
- **(2) Same source elsewhere? No second venue found.**
  - The 204 kit deployers plus the funder contracts have never launched on v4 here (0 of 171,014 v4 launches in the day files; 0 of 91,047 in crew.db).
  - None of them has any activity on Base, Arbitrum or BSC.
  - Only the farm funder below has 53 transactions on Ethereum mainnet.
- **(3) rep1_pre on the wide cell: drop it, because it costs money.**
  - All 70 wide trades: **+$2,480 a day** (A +$2,863, B +$2,183).
  - Skipping coins with more than 5% repeat ETH keeps 54 trades at a better average (2.29x vs 2.13x) but only **+$2,178 a day** (A +$2,320, B +$2,068).
  - The 16 skipped coins still averaged 1.60x (5 of them hit the 4.8x target), so skipping them throws away about $300 a day. Cuts of 10% and 20% lose money too.
- **(4) The near-equal 1.4 to 1.5 ETH wallets are a proven farm, but there is no skip rule worth using.**
  - 11 of the 16 biggest repeat wallets in the 21 to 24 Sep coins got their first ETH as exactly 0.002 ETH, inside about 3 s (blocks 69,614,623 to 69,614,656), each from a different wallet. 4 more got their first ETH together in one transaction (block 65,680,181).
  - The 5 senders I traced were all funded together in one block (69,614,401) by `0x3393cab3fcfc5668b46d83b89f924c01f00aa766`.
  - The same `0x3393cab3fcfc5668b46d83b89f924c01f00aa766` also funded Kit A deployers (for example `0xf244ad0588c0a629af456162fb47d4976bf64b87`, `0x6e7aa62c9c7da7d661d731ac537c6e57d41625a0`, `0xeb21bd9e4fca69c0b4c102fc8a03e0158c63f50a`, through `0x03802649e469ed5680c58dff9e65425e22d21cc5`).
  - **So one operator launched Kit A coins and bought them pre-gate with its own wallet tree.**
  - The 7 coins it pumped still averaged about 1.46x on our rule (2 hit 4.8x), so skipping them loses money.
  - Coder: tag `farm_0x3393` and `kit_A` / `kit_B` on every candidate for tracking only. Don't skip on them.

## Method
- **Pools:** every Uniswap v3 PoolCreated event with WETH since block 61,500,000 (7,108 pools, all fee tiers), from factory `0x1f7d7550b1b028f7571e69a784071f0205fd2efa`.
- **Launch source:** the create transaction's sender (deployer) and "to". Kits were grouped by the keccak hash of the "to" contract's bytecode.
- **Kit creators:** binary search on eth_getCode found each copy's creation block. In 8 of 8 checked, the launch sender deployed it in a direct create.
- **Gate coins:** the 114-coin $1M gate set; wide rule = Analyst 2's 70 trades (`sw-a2-16-wide-trades.csv`).
- **First funders:** binary search on balance, then the call trace of the funding block.
- **rep1_pre:** pre-gate buyers followed through routers. Memory = earlier wide-gated coins only. Contracts removed. $ a day = $500 × (x − 1) summed per day over 16 days.

## Limits
- The bundle-buy share is from 4 Kit B coins only.
- Kit B callers were not traced further back. One operator vs a shared tool is open.
- Other chains were checked only by wallet activity, not by bytecode search.

## Asks for Supervisor
- Coder: keep the wide rule as is. **Do not add the rep1_pre skip** (−$302 a day). Add tags `kit_A`, `kit_B`, `farm_0x3393` and the Kit B bundle share for tracking.
- PP: add a live counter of Kit B launches per day (bytecode hash 89885b75fd29, funder `0x84157d3d24630c297849c84d49ae7f1805c1b60f`). If it drops to zero, our coin flow roughly halves; alert the Supervisor.
- Virality Expert: Kit B coins carry a 26% to 38% dev bundle before the gate, so "sellable $1M" partly counts the dev's own ETH there. Worth a hold-side check.

First principles score: 85%. The source comes from each pool's own create transaction and the bytecode hash. Funding comes from call traces. Money comes from the rule's own trades. Missing: the full funder tree behind the Kit B callers, and a bytecode search on other chains.

Scripts: /root/a1/a114/src.py, an.py, creator.py, fund.py, fsrc.py, rep.py, farm.py, farm2.py (review container). Latest: http://159.223.88.168/hx-c9ois/sw/a1-14-v3-source.md