shftkxns…Vikj

blueshift

Risk score

Low risk

Total extracted

27.97SOL

Sandwich involvement

5.1%

Wide sandwich rate

5.1%

Avg per slot

0.002727201SOL

Engine evidence

Every metric below maps 1:1 to a field the detector emits on each attack.

Atomic bundle

0.0%

0 / 50 attacks

Attacker paid for atomicity in this leader's slot — strongest direct involvement signal.

Same-block

10.0%

5 of 50 via header

Front+victim+back in one slot — only possible when this validator is leader.

AMM-replay coverage

92.0%

46 replays · 4 fallback

Loss computed from pool math, not amount_out heuristic. Carries lower/upper bounds in raw output.

Counterfactual flips

of 50 flagged

Rule-based said sandwich; replay says attacker took a loss. Shown so you can audit, not hidden.

Replay-derived loss

1.323SOL

interval [1.121 – 1.524] SOL

Sum across 45 of 50 attacks where AMM replay returned per-step parser/model residual bounds. The remaining 5 fell back to point estimates.

Loss source breakdown

raydium_v4 BYZVLgA4…zFR588

raydium_v4 HpLgk1ee…pz6G62

raydium_v4 GvVMAy6F…BShg46

raydium_v4 DbAQNghs…zJW143

raydium_v4 JAqU6ctx…5vtg42

Latest 20 events on this validator.

Time	Type	Pool	Victim TX	Bundle	Detection	Replay	Loss (SOL)	Slot	Severity
05/13, 01:26 AM	Wide sandwich	pump_fun ArbfcD…	tmfxeMdKaX…	Organic	Cross-slot	AMM	0.0000	419,370,870	low
05/13, 01:26 AM	Wide sandwich	pump_fun ArbfcD…	5vzqB2LzFr…	Organic