Meteora LP — Admin

Equity — money over time — (wallet + open positions) − net funding; the only complete P&L

Performance by profile — per-trade estimate, closed only (incomplete: pooled-sweep recoveries & rent land at the equity level, not per trade — see Equity above)

Trade P&L by window — per-trade, incomplete

Closes by reason

Open positions by profile

Open positions — real (деньги) и paper (виртуальный двойник) помечены бейджем

Cumulative realized P&L (SOL)

Realized P&L distribution (%)

By profile

By close reason

Closed positions — click a row for candles + decision records

Score components (top candidates)

Candidates — what we monitor & why not entered

Learned overrides — clamped adaptive layer over scan_profiles; disable a row to revert that param to operator config

Range width → outcome (learned) — per pool-class; the range learner's evidence, gate before enabling auto-apply

Shadow vs real (closed) — per-variant virtual books vs the live book; the other pre-flight gate

Post-exit outcomes — что пул делал ПОСЛЕ нашего выхода (окно 24ч)

Intelligence — agent recommendations on closed positions, embargo-gated · advisory-only · actionable=false

Pool memory (win-rate by pool)

Recent outcomes

Lessons

Decision log — click a row for the full record

Missed winners — gates that may be too tight (loosen to admit +EV pools)

Avoidable losses — features that separate losers (tighten/add a gate)

Agreement — replay findings vs REAL borderline-shadow outcomes (near-miss pools opened as shadows)

Recovery — hold vs sell on losing exits: shadow BidAsk ladder above exit price, baseline (sold immediately) = 0%

Run a backtest — исторические SKIP/ENTER решения переигрываются с новыми параметрами (пустое поле = текущее значение профиля)

Profile From To Range low % Range high % Grace down, min Grace up, min Gate overrides Sweep Regime

Result

Recent runs

Execution — read-only, change via SSH

Equity — money over time — (wallet + open positions) − net funding; the only complete P&L

Performance by profile — per-trade estimate, closed only (incomplete: pooled-sweep recoveries & rent land at the equity level, not per trade — see Equity above)

Trade P&L by window — per-trade, incomplete

Closes by reason

Open positions by profile

Open positions — real (деньги) и paper (виртуальный двойник) помечены бейджем

Position detail

Pool trajectory — TVL / volume / fee-rate (% of entry)

Cumulative realized P&L (SOL)

Realized P&L distribution (%)

By profile

By close reason

Closed positions — click a row for candles + decision records

Position detail

Pool trajectory — TVL / volume / fee-rate (% of entry)

Score components (top candidates)

Candidates — what we monitor & why not entered

Why not entered

Learned overrides — clamped adaptive layer over scan_profiles; disable a row to revert that param to operator config

Range width → outcome (learned) — per pool-class; the range learner's evidence, gate before enabling auto-apply

Shadow vs real (closed) — per-variant virtual books vs the live book; the other pre-flight gate

Post-exit outcomes — что пул делал ПОСЛЕ нашего выхода (окно 24ч)

Intelligence — agent recommendations on closed positions, embargo-gated · advisory-only · actionable=false

Pool memory (win-rate by pool)

Recent outcomes

Lessons

Decision log — click a row for the full record

Decision record

Missed winners — gates that may be too tight (loosen to admit +EV pools)

Avoidable losses — features that separate losers (tighten/add a gate)

Agreement — replay findings vs REAL borderline-shadow outcomes (near-miss pools opened as shadows)

Recovery — hold vs sell on losing exits: shadow BidAsk ladder above exit price, baseline (sold immediately) = 0%

Run a backtest — исторические SKIP/ENTER решения переигрываются с новыми параметрами (пустое поле = текущее значение профиля)

Result

Recent runs

Execution — read-only, change via SSH

⬡ Meteora LP — sign in

Equity — money over time — (wallet + open positions) − net funding; the only complete P&L

Performance by profile — per-trade estimate, closed only (incomplete: pooled-sweep recoveries & rent land at the equity level, not per trade — see Equity above)

Trade P&L by window — per-trade, incomplete

Closes by reason

Open positions by profile

Open positions — real (деньги) и paper (виртуальный двойник) помечены бейджем

Position detail

Pool trajectory — TVL / volume / fee-rate (% of entry)

Cumulative realized P&L (SOL)

Realized P&L distribution (%)

By profile

By close reason

Closed positions — click a row for candles + decision records

Position detail

Pool trajectory — TVL / volume / fee-rate (% of entry)

Score components (top candidates)

Candidates — what we monitor & why not entered

Why not entered

Learned overrides — clamped adaptive layer over scan_profiles; disable a row to revert that param to operator config

Range width → outcome (learned) — per pool-class; the range learner's evidence, gate before enabling auto-apply

Shadow vs real (closed) — per-variant virtual books vs the live book; the other pre-flight gate

Post-exit outcomes — что пул делал ПОСЛЕ нашего выхода (окно 24ч)

Intelligence — agent recommendations on closed positions, embargo-gated · advisory-only · actionable=false

Pool memory (win-rate by pool)

Recent outcomes

Lessons

Decision log — click a row for the full record

Decision record

Missed winners — gates that may be too tight (loosen to admit +EV pools)

Avoidable losses — features that separate losers (tighten/add a gate)

Agreement — replay findings vs REAL borderline-shadow outcomes (near-miss pools opened as shadows)

Recovery — hold vs sell on losing exits: shadow BidAsk ladder above exit price, baseline (sold immediately) = 0%

Run a backtest — исторические SKIP/ENTER решения переигрываются с новыми параметрами (пустое поле = текущее значение профиля)

Result

Recent runs

Execution — read-only, change via SSH