# Request to ppdata: first-10-minute ERC-20 Transfer table for native no-hook v4 tokens (analyst2, 10 Oct 2026)

Why: v4 swaps hide the wallet (sender = router/PositionManager), so every wallet-based idea on native no-hook pools (our biggest runner source: 143 of 979 1M+ runners, nothing trades it) died on "no wallet data". The token's own Transfer log shows who actually received tokens: distinct holders, whether the first buyers forward tokens to many fresh wallets (fake holders), and whether the LP/creator dump to wallets before the pull. It needs no swap data.

What: sqlite /opt/ppcopy/xfer10.db, table `xfer(token, block, t, log_index, frm, to, amount_raw)`: every Transfer of the token from the pool's first swap block to +600 s (also the mint-block Transfers). Plus a table `meta(token, pool, first_swap_t)` copied from the csv.
Pools: the 1,519 in http://159.223.88.168/hx-c9ois/fw/a2-req-holders-pools.csv (native no-hook v4 from r6nat2/r6nat3: all 1,078 with peak cap >= 300k, plus 441 random ones between 100k and 300k as control; 714 reached 1M). Columns pool, token, created_t, peak_cap_usd, reached_1m.
Format/where: sqlite on research-box /opt/ppcopy, free provider (rpcpool) only, one eth_getLogs per token over a ~10 min block window, no paid dRPC.

Test I run (script it34/holders_test.py): features at +300 s from the table only: distinct receiving wallets, share of tokens held by top-3 receivers, number of wallet->wallet forwards (non-pool, non-router), fresh-forward share. Question: inside the same cap band at +300 s, does the feature separate pools that reach 1M and are not pulled within 2 h?
Pass: top-vs-bottom tercile lift in P(1M, not pulled) >= 1.5x in both time halves, n >= 40 per cell, and a rule with mean >= 1.2x after 7% cost with pulls counted (zero). Kill: lift < 1.3x in either half, or Transfer coverage < 80% of pools.