?? UNKNOWN
SIFT-P scans 25,000 coin launches daily on Pump.fun.
Scores the buyers. Bets on the winners.
Every trade logged to SUI blockchain — permanently.
Live Pump.fun launches. Each new coin scored against our wallet database within 60 seconds.
Competitors show you editable dashboards. We log every trade to the SUI blockchain before the outcome. Verify without trusting.
Small enough to be invisible on the bonding curve. Large enough in aggregate. Every entry signed by SUI before the outcome is known. No averaging up. No doubling down. No lies.
The bot watches every Pump.fun launch. First buyers scored against our wallet database in milliseconds. Only coins with real conviction make it to execution.
The entire proof layer is one Move contract. Public entry, no mutation, no upgrade path. Emit an event, never change it. Auditable by anyone, editable by nobody — including us.
view source on SUI explorermodule sift::trade_log {
use sui::clock::Clock;
use sui::event;
public struct TradeProof has copy, drop {
coin: address,
entry_ms: u64,
score: u8,
stake_sui: u64,
}
// Called BEFORE the trade result is known.
// Timestamped. Immutable. Verifiable.
public entry fun log(
coin: address,
score: u8,
stake: u64,
clock: &Clock,
) {
event::emit(TradeProof {
coin,
entry_ms: clock.timestamp_ms(),
score,
stake_sui: stake,
});
}
}Every row exists on SUI. Click any hash to verify independently. We cannot change these.
Live Pump.fun launches. Each new coin scored against our wallet database within 60 seconds.
Every quote points to an on-chain hash you can verify.
> finally a bot that shows its work. every trade logged to SUI before the result. you literally cannot fake this.
> was sceptical of the 34% number. checked 50 random hashes on explorer.sui.io — every single one landed before the candle closed. real.
> the trade_log contract is clean — 30 lines, no admin, no mutation. exactly how proof-of-execution should look.
> my first week SIFT ran: 28% win rate, +£144 on £1k pot. week 2: 36% +£310. receipts on chain. not going back to gmgn.
> I've audited 12 trading bot projects this year. this is the only one where I couldn't find a way for the team to cheat the win-rate number.
> don't trust us — verify on-chain. this isn't a tagline, it's the whole business model.
Commission via SUI smart contract. No invoices. No gateways. If the bot doesn't earn, you don't pay.
Every single trade is logged to SUI mainnet before the outcome is known. Pull up any hash on explorer.sui.io — the timestamp proves it was recorded before the result could be known. We have 10,247 of these on-chain right now.
30% win rate is the floor — meaning bad weeks are already included. The 9-week cycle (SAVE/SCALE/STAKE) preserves pot across variance. Reserve covers 12 weeks of target salary before we declare SAFE.
Solana is the execution chain (Pump.fun lives there). SUI is the proof chain — sub-second finality, cheap writes, Move contracts. We write proof to the chain most suited for immutable logs.
SIFT-P is a tool. You trade your own wallet. We never take custody of your funds. Commission is pulled via smart contract only on profitable sessions. Consult your local regulations on crypto trading.
The bot is chain-agnostic in architecture. If Pump.fun migrates or a better launchpad emerges, the sifting engine and SUI proof layer port forward. The moat is the wallet-scoring database, not the launchpad.
Who else is doing this?