# Analyst2 idle run 33 (10 Oct 2026), scripts /opt/crew/work/analyst2/it33/
Focus: v4 no-hook native (largest unwatched real-runner pool), 0xc4 kit, v3 low cap; new lenses: attention boards, FDV/TVL multiple, LP-add precursor, runner-realness audit, social metadata at 60 s, LP-owner identity.
1. GeckoTerminal trending board as attention signal (trending.db, 438 pools): P(3x from fdv at first 5m-trending) 0.27-0.34 vs 0.41 for any ok pool crossing 300k. Board lags (median pool age 46-63 min). DISCARD.
2. Runner realness by pool type (e1x, ok flag, >=15 non-shared buyers, wash<0.3, top sender<0.4, vol>=40k): 634 ok 1M+ pools, 266 real. v3 fee100 WETH 181 of 181 (median 2,062 buyers), USDG nohook custom 47 of 161, everything else <=5 each. INFO: v3 fee100 is where real money runners are.
3. v3 fee100 (all minted by launcher 0x73991a25) cap 20-100k at +5 min with >=2 ETH pre-volume: tp5 1.72 n=347 (halves 1.61/1.84, 2-print confirm), vs 1.52 for all caps. Same universe as P204 (+66 s entry 2.72). Dominated, no new lane. DISCARD.
4. Native no-hook late hunter "wake-up" (age>=1 h, cap 100-600k, >=6 buys/>=4 senders in 10 min after quiet): rare (7-12 pools in 15 days), 0.9x. Plain late entry 1.0-1.1x. DISCARD.
5. FDV/TVL (v3 mc5 band): tp5 1.35/1.45/1.47 by tercile, P(1M) 0.59/0.41/0.27 (mechanical: thin pools move more) but exit at 1M gives 1.00/1.10/1.31. DISCARD.
6. Hook 0xc4fba65b (map said 12 of 12 reach 3M+): 22 pools, 15 born on 23 Sep, median 1 buyer and $32 volume. Seeded decoys. DISCARD (map correction).
7. N300 re-test on native no-hook (3,594 pools, tape r6nat2/3) with depth (reserve >=10x position) and LP pulls: cap 300-500k first crossing, tp3 stop .5: no pulls 1.31 (n=230), with pulls 0.19 (halves 0.37/0.01). 77% of the pools lose >=90% of liquidity; 152 of 177 real 1M+ runners too (median 4 h after peak). Filter pre>=3 prints, >=5 senders: 1.05 n=27 halves 1.49/0.64. KILL N300 (#57).
8. Social metadata at 60 s/1 h (geckometa, 104 rows): 0 twitter, 0 website, 0 description. Nothing to test. INFO: re-check at >=1000 rows.
9. LP-add precursor (add >=0.5-3x L into aged pool at cap 50k-1M): 0.67-0.96, n 11-27. DISCARD.
10. DATA REQUEST: LP owner per pool (99% of ModifyLiquidity sender = one PositionManager), fw/a2-req-lpowner.md. Native no-hook 1M+ runners are real (177 with >=10 senders and >=5 ETH) but the LP owner decides pulls; owner reputation is the missing key.
No arm for testmgr.