# A1: V3-005 template: v3 fee-100 pool that opens with an exact 0.05 ETH first buy (9 Oct 2026, analyst1)
How found: grouped all 8,329 WETH v3/v4 pools in r6weth by creation template (first liquidity ticks/amount, liquidity L), not by hook or fee. Most templates are noise; one v3 template has 2% total loss.
Template (tape): v3, fee 100, no hook, position-manager 0x73991a25 mint of 3.333e26 tokens with L=1.227e22, then the first swap is a router buy of exactly 0.05 ETH. 107 pools 21 Sep - 7 Oct (about 6/day, steady, none missing days except 30 Sep).
Rule (frozen):
- Detect: v3 WETH pool, fee 100, first swap is a BUY of 0.05 ETH (+-0.0005) through the router. (In v3live this is observable; the mint amounts are not stored there.)
- Entry: first swap after which the pool ETH reserve is >=0.2 ETH (swap-flow reserve) within 60 s of the first swap. Buy 0.02-0.05 ETH.
- Exit: sell at 5x (pool price vs entry price); otherwise sell at 20 min; if the LP is pulled before, count 0.
- Costs included: 1% fee each way, 2% position impact, 3 s lag.
Evidence:
- Tape (in-sample n=69-101 depending on gate): tp3 2.5, tp5 3.8, tp10 5.2-5.8; halves 2.57/2.47 (tp3); 3% pulled to zero. Entry delay curve: +5 s tp5 3.79, +60 s 3.86, +300 s 2.18 (be in within 60 s).
- Forward OOS in v3live (7 Oct 19:37 - 9 Oct 14:20 UTC, first swap exactly 0.05 ETH, enterable by the same reserve gate): delay 10 s n=11 tp3 2.59 tp5 4.23; delay 60 s n=20 tp3 2.48 tp5 3.93, 10% zero; delay 300 s 2.08/2.17. Consistent but n=20.
WARNING: in v3live 274 pools have a 0.05 ETH first swap (24/36/187/27 per day) and 99% of them are pulled with median max price 1.04x: a dead spam population shares that first swap. The reserve>=0.2 ETH gate (real-time, observable) is what separates them; without the gate this rule loses. The gate leaves roughly 3-10 pools/day.
Open: outside money unknown (tx.from only partly known); same operator as TINYV3 / kit? not checked; n=20 forward. Fits the paper test as a small arm. Scripts: research-box /opt/crew/work/analyst1/p7-p11.py, p3.py.