# A1 #1885 rug filter (hook/LP owner) - 11 Oct 2026
Verdict: NO FILTER. Nothing pre-entry separates the 5 rugs from the other 57 live mc5 trades.

1. The mc5 pools are Uniswap-v3 style (20-byte pool address, fee100), not v4. There is no hook to check (the hook/sell-revert honeypot exec found on 0x7c698f/0x5ebda7/0x7262b2 is the v4 fml2 picks, a different bot).
2. Tested the LP side for all 57 mc5 buys with on-chain data (r2/lp.py, f.py: first Mint tx per pool, tx.from, LP NFT recipient, code, nonce): 5 rugs (0x313e89, 0xd012ab, 0xdc5400, 0x9fba63, 0x2d8532) vs 52 others (4 winners 13-29x).
   - Every rug has its own deployer and its own LP recipient: no repeat operator across rugs (so no blacklist).
   - LP recipient is an EOA with nonce 27-91 in all 5 rugs; same range as winners (7-143) and stops (0-195). Not a contract/locker in any rug. No from==to pattern. Single mint everywhere.
   - Cap at entry (263k-712k) and pool WETH (9.3-15.3) overlap the rest completely.
3. Money: a rug costs ~(0.5 - 0.005) x size over a normal stop; 5 in 62 trades = about 0.03 ETH of the 0.051 headline, and the stop already caps the other 45 losers at ~0.5. Any filter needs to drop a 13-29x runner less than ~1 in 25 candidates; with n=5 and no feature, no such filter can be shown.
4. Execution cannot help (exec confirmed one-step pull). Only structural option left: smaller size on entry, but that cuts the winners equally. Not recommended.
5. Open lead only: the dump-in-one-block seller wallets (0x0d1c64af.. in two rugs/stops) from tm-rugs-mc5; needs pre-entry visibility of a seller, which we do not have at +5 min. Parked.