# #2139 Rug gate replay on the wlink rugs (research, 12 Oct)
Script rug2/wl.py (creator = tx.from of first pool log and first token log, free RPC). Seed list = the 12 wallets in fw/exec-rugs-12oct.md.

| pool | entry (BKK) | cap | pool-create sender = deployer | on seed list? | gate v1 result |
|---|---|---|---|---|---|
| 0x8a906b | 14:44:54 | $45k | 0xa858ac99 | no | PASS (first rug, nothing to know) |
| 0xd662a5 | 14:49:49 | $49k | 0x3e67ddda | no | PASS (buy_fail, no money lost) |
| 0xc51c1e | 15:02:35 (+15:03:15 on wlink1) | $50k/$72k | 0xaf3821a6 | no | PASS: only 1 pull in the last 2 h, pause needs 2 |
| 0x610c7e | 15:05 | - | 0xbe59d50d | no | BLOCKED by "paused: 2 pulls in 2 h" (correct if rug, outcome unknown) |

Findings
1. Wallet list: 0 of 3 caught. All creators are new singleton wallets (create sender = deployer), no history. Same conclusion as #2131/#2134: wallet history cannot stop first-timers.
2. Pause rule: as set (2 pulls) it fired only after the 2nd rug. The 3rd entry (0xc51c1e) came 18 min after the 1st pull and 13 min after the 2nd buy; with N=1 for wlink it would have been blocked (and 0xd662a5 too). Cost: wlink pulls are full losses (sells fail, ~0x, no 0.5x stop), unlike mc5 where the stop caps ~0.5x, so for wlink pause-after-1-pull is cheap. I have no wlink winner data to price the pause; wlink has 0 winning live trades so far.
3. wlink1 and wlink entered the same pool 40 s apart: both lanes lost on 0xc51c1e (2 x $10). A shared gate/pool-dedupe across lanes would halve that exposure.
4. Common fingerprint of the 3: cap $45-72k at ~100-150 s after creation, pool WETH 4.4-6.7 ETH (tiny pools, ~$10-15k). Not testable as a rule on 3 cases; ask ppdata for a wlink replay set (all wlink signals since 6 Oct with LP-pull outcome) before ruling.

Recommendation (exec): for wlink/wlink1 set PULLS_N=1 (2 h pause) and share the pause across both lanes; keep enforce. rug_gate flag in registry: the gate demonstrably does NOT stop first-time rugs; only the pause limits the damage. Treat the wlink lane as high rug risk until n>=20 live trades say otherwise (Ben's call).