Live trade log — HYPE-PERP · Hyperliquid mainnet

HYPE Scalper

Every closed trade from an AI-operated perp bot, published as executed on-chain. Two-sided (long/short), signal-driven, no manual entries in this log. Period: . All figures are percentages of position notional — capital size is intentionally not disclosed.

HYPE mark

Position

Loading

Entry

Open ROI

Connecting to Hyperliquid…

Cumulative ROI

sum of per-trade returns

Trades

Win rate

Avg win / loss

Median hold

Cumulative ROI — running sum, %

Closed trades

#DateSideEntry → ExitHoldTrigger in → outTrade ROICumulative

Methodology

Trade ROI = realized P&L of each closed position ÷ position notional at entry (before leverage). Cumulative is the arithmetic running sum, not compounded. Trades are reconstructed from Hyperliquid fill records; partial exits within the same position are merged. Fees are excluded per-trade and total <0.4% across the period.

The bot trades one market (HYPE-PERP) using public data only: order-flow delta, whale TWAP detection, top-account positioning, funding/OI dynamics, and volatility-regime gating. Longs to date: · Shorts: — the short side is newer and still being tuned. Losses are shown unedited.

Trigger codes

signal composite entry score crossed its threshold (CVD + OI + funding + trend + book) · fast 1-minute velocity spike with flow confirmation · flip reversal detected — position cut and direction switched · fast-flip sharp counter-move while in position — instant cover & reverse · confirm breakdown confirmed below EMA after a flip signal · tp1 / tp2 / tp take-profit levels · trail trailing stop from peak · stop hard stop-loss · dca scale-in add · manual / limit operator command. Sub-minute holds are fast-reaction scalps; multi-day holds are trend runners — the bot does both.

This is a small experimental system published for transparency, not investment advice or a solicitation. Past performance does not predict future results.

@Not_A_De_Gen profile picture

Built by @Not_A_De_Gen

Building this bot in public — signals, losses, and architecture breakdowns as they happen.

DATA: HYPERLIQUID PUBLIC API · UPDATED