# fmmc5 live: rugged pools and "peak 1.0x" trades (Test Manager, 8 Oct 2026 ~20:45 Bangkok)

For Analyst 1, item A1-RUG. Source: firstmin /opt/fm/mc5.db (read-only) plus the public Robinhood RPC. Nothing on firstmin was changed.

**Kit:** none. fm_mc5.py has no kit tagging; every live trade is the plain mc5 rule (0.5x stop, 4 h, cap 250k-1M, entry +5 min).

## Rugged pools (sold at ~0.01x)

| buy (Bangkok) | size | pool | token | buy tx | sell tx | cap at entry | pool WETH at entry | mints / burns in mc5.db |
|---|---|---|---|---|---|---|---|---|
| 8 Oct 20:05:32 | $50 | 0x07b59c1e7db8faec5ba64077a5c93abdd32cc90d | 0xd012847dce479c39a4237adb17c461ab7c9eb54a | 0x4d2fecdf29afb8d3a95e6659d673cd348cd8c28b7f3507f78b3d0d8c280da15e | 0x01880e1cd6e01997eb59bc81b735dd6902afe28407c94d2d08402dd9412818fc | $390k | 11.43 | 1 / 2 |
| 8 Oct 20:12:43 | $50 | 0xd002fed10b877209682658bf6ba8d81fb468745a | 0xdc5400cae8c82bbac1bc96dbec38f48a5fdb5610 | 0xb9522303ee6149f3c81836c4cb6a0fe9c557f26173b24506a8db458afc7394db | 0xf5201a14481df0804be734795478a5ea1cbe74d40aeca733a5ead675c658ecd7 | $336k | 10.89 | 1 / 2 |
| 7 Oct 02:38:31 | $10 | 0x4fc807391d8dd76992043950248b26a610231075 | 0x313e89b7798d0ec8b755fd46a9da9a14a4ece708 | 0x89da0e7097e8a2fea2b1287f38262acb0d7cd9b122f0d1e4a416e0d8a89272aa | 0x2cf3c8f72295d561ee72ba0e453b64ad28e673bba0cc5dc3ff0c62fe3a08a5bf | $712k | 15.28 | 1 / 1 |

Caveat: I have not confirmed whether the mints / burns counters in mc5.db are counted at entry or up to exit. Check against the tape before using them as an entry-time feature.

On both 8 Oct rugs the stop fired 1.2-1.8 s after the cross, but pool WETH was already 0.0002 / 0.0031. GMGN refused the sell (GEvmValueLossTooHigh), and the 60 s retry sold for dust.

## Second pattern: 5 of 10 trades today never went above our buy (peak 1.00x)

Peak 1.00x trades: 16:55, 18:33, 18:44, 19:30, 20:05. On 6-7 Oct, no trade had peak 1.00x (lowest 1.20x). These stopped out 10-62 s after the buy.

On-chain swaps between our buy and our sell (public RPC) show the same shape in each one: one or two wallets sell in ~1.0 WETH chunks every few blocks, starting right after our buy, while fresh ~0.3-0.5 WETH buyers keep coming in.
- 16:55 pool: sellers 0xf45d2cb2a0…, 0x054e840726…, 0x1e613d1ec9… (each 2-5 sells of ~1.0 WETH).
- 18:33 and 18:44 pools: **the same seller wallet, 0x23736431f8…**, does 5 and 8 sells of ~1.0 WETH, ten minutes apart, in two different coins. A second wallet, 0x0d1c64af20…, joins in the 18:44 pool.
- Small 0.0039 WETH buy-then-sell wallets (0xbc7dd06a14…, 0x6949736f36…, 0x573a13e752…) show up in several pools.

The same wallet distributing into buyers across different coins suggests one operator (deployer or bundle) selling into the +5 min buyers, which matches your deployer hypothesis. Wallets are shown as prefixes; the full addresses are on the tape (pool + the block ranges our buy and sell txs give).

## Separate, for the RPC owner
Paid dRPC refused with HTTP 403 ("billing or cap") from about 16:19 to 17:01 Bangkok; the feed fell back to the public node.