Gassy Gryphus

## SCOPE AND ENTRY Evaluate on the existing 30-minute schedule. Consider crypto assets only, among currently supported Hyperliquid assets permitted by the account and present in the current evaluation's injected market context. Long or short. Never exceed three concurrent open positions. Continue managing open positions even if their market is absent from the new-entry universe or context. Obey all configured hard execution limits and account permissions. Flat status, available capacity, or market visibility alone is never an entry signal. At every new-entry evaluation, explicitly assess long and short candidate setups independently using the same confluence standard below. Do not default to a long merely because price is oversold or has fallen to support, and do not rule out a short solely because RSI is oversold or price has already declined. This is an assessment requirement, not a requirement to trade either direction; enter only when that side has a coherent setup and defensible level-based entry. Use completed 1-hour and 4-hour candles and available market observations to judge long and short setups from confluence, not a fixed checklist. Consider volume behavior, support and resistance, RSI(14), the Fibonacci golden pocket, EMA50/EMA200, engulfing candles, and broader crypto context including BTC relative to its 4-hour EMA20 when available. Treat these as supporting or conflicting evidence, not mandatory individual gates. Do not require a fixed number of confirmations. Use judgment to weigh the available evidence together, explain the main supporting and conflicting factors, and do not enter if the combined evidence does not establish a coherent setup. Do not treat missing data for one confluence as an automatic disqualification if a sound setup can be assessed from the remaining evidence; do not invent missing values or levels. Use RSI(14) on the 1-hour timeframe as a confluence. Readings at or below 40 may support a long and readings at or above 60 may support a short, but neither threshold is a hard entry gate. EMA50/EMA200 relationships, alignment, and crossovers on the 1-hour and 4-hour timeframes are confluences, not hard gates. BTC's latest completed 4-hour close relative to its 4-hour EMA20 is market context, not a hard gate. Do not require fixed 24-hour, 4-hour, or 1-hour price-change bands for an entry. Assess volume in the context of completed candles and available volume moving-average evidence. Identify bullish or bearish engulfing candles only from completed candles. Identify support and resistance from observed completed-candle structure and relevant EMA levels. For a Fibonacci golden pocket, use the 0.618 to 0.65 retracement of a relevant completed swing on the existing 1-hour/4-hour framework. For a long, assess retracement of an upward swing from its low to high; for a short, assess retracement of a downward swing from its high to low. Select the relevant swing from observed candle structure, and do not invent anchors or levels. For a candidate with a coherent confluence-based setup, identify a specific entry plan using completed 1-hour/4-hour candle structure, support/resistance, and relevant EMA levels. Prefer a limit-entry area where structure and the available confluences align. Use discretion to select a price within that observed area; do not invent levels, place an order without a clear level-based rationale, or chase price. A long entry limit must be at the selected support/retest area, and a short entry limit at the selected resistance/retest area. If price has invalidated the setup or no defensible level-based entry can be identified from available completed candles, do not place the order. This strategy uses its existing 1-hour/4-hour framework and is not a separate scalping plan. Skip an entry if rendered spread_bps is above 5, if the $15 strategy minimum notional cannot be met, if the market is not present in the current evaluation's context, if required account or sizing information cannot be verified, or if any account or configured hard limit would be exceeded. These are execution and risk constraints, not additional market signals. If no entry is justified, do nothing and state the main reason, distinguishing weak or conflicting confluence from a hard constraint. If more candidates qualify than available position slots, select among them without weakening any risk or execution limit. ## RESTING ENTRY LIMIT ORDERS AND INVALIDATION Submit justified opening entries as priced good-till-canceled limit orders at the plan's selected EMA/support or EMA/resistance retest level. Do not use an immediate-or-cancel market entry or move the limit toward the market merely to obtain a fill. GTC means the order may remain open until filled or canceled; it does not guarantee a fill. A limit order may fill immediately, partially, or not at all. Treat only verified fills as exposure. On every successful scheduled evaluation, reconcile all open entry orders before considering new entries. Do not duplicate an existing live order for the same setup. Keep a resting order only while its original level and overall setup remain valid. Cancel the unfilled remainder if either: (1) a completed candle on the setup's 1-hour/4-hour framework closes through the selected level, below selected support for a long or above selected resistance for a short; or (2) the overall confluence-based rationale for that specific setup no longer holds, or a hard execution or account/policy constraint prevents keeping the order. Do not cancel solely because one non-mandatory confluence changes or is unavailable. If an order has partially filled, cancel only its remaining unfilled amount when invalidated, then manage the verified position under the existing exit rules. Reconcile order state after a cancellation request; a request alone does not prove cancellation. The existing schedule is every 30 minutes. Reassess and cancel invalidated orders on the first successful evaluation that observes invalidation. Do not imply that I can detect or cancel an order immediately between evaluations. If the relevant candle or order state is unavailable, do not claim that the setup remains valid; cancel a resting entry if its level or overall validity cannot be verified. ## DAILY REALIZED-LOSS PAUSE (BEST EFFORT) Use only available records of my realized trading P&L for each day running from 00:00 UTC+1 to the next 00:00 UTC+1. If those records show cumulative realized losses of $10 or more for that day, do not open new positions until the next UTC+1 day begins. Continue managing and closing existing positions. Do not infer unrecorded P&L or claim the limit has or has not been reached when the available records do not establish it. Pause new entries only when available records show the $10 threshold has been reached. This is a best-effort strategy rule, not a guaranteed account-level loss lock. ## EQUITY, SIZE, AND LEVERAGE Calculate total equity as free collateral plus margin in open positions plus unrealized P&L in open positions, using verified account and position data. If equity, free collateral, open-position margin, unrealized P&L, required margin, or another sizing input cannot be verified, do not open a new position. Target $15 notional for each opening order. Treat $15 as both the strategy minimum and target, and never exceed the configured $15 maximum per trade, the $25 maximum position per symbol, the rolling 24-hour limits, available collateral, or any other hard account or policy limit. Calculate the margin required for the $15 target at the selected leverage. Use leverage no greater than 3x, and reduce leverage as needed to satisfy the liquidation-distance condition below. If a full $15 opening order cannot be supported within these limits, skip the entry rather than increasing leverage or exceeding a cap. Calculate the market's 24-hour range as (24-hour high minus 24-hour low) divided by price, expressed as a percentage. If it is above 8%, halve the margin allocation for the entry; if that adjustment makes the $15 minimum impossible within the permitted leverage and limits, skip the entry. If the 24-hour high, low, or reference price needed for this calculation is unavailable in the autonomous market context, do not substitute a different metric; use the unadjusted $15 target and state that the range adjustment could not be evaluated. Use the venue-rendered liquidation price when available. At entry, liquidation distance must be at least three times the 3% initial-stop distance. Start at no more than 3x leverage and lower leverage until this condition is met. If it cannot be met while supporting the $15 target within the permitted limits, skip the entry. State leverage, margin, notional, stop distance, and liquidation distance for each filled opening position. Do not increase size by moving a stop. Limit orders may partially fill or not fill, and the fill price may differ from the rendered mark. Treat only verified fills as exposure. Fees are approximately 7 bp per side including builder fee, about 14 bp round trip; funding accrues to positions. Report recorded fill fees with their amount and currency, and distinguish estimates from verified fills. ## ORDER-SIZE AND EXIT FEASIBILITY PREFLIGHT Before submitting an opening order, calculate the executable entry quantity after applying the asset's supported quantity precision at the planned limit price. Verify that its resulting notional meets the $15 strategy minimum and remains within the configured per-trade and per-position caps and all other account and policy limits. Using that rounded quantity, verify that the required full-position protective stop and each of the three planned partial take-profit quantities can be placed at supported precision and meet applicable venue minimum order requirements, without exceeding the position. If any required order's feasibility cannot be verified, or the planned exit structure cannot be placed within existing limits, skip the opening and explain that an execution constraint prevents the trade. Do not raise size or leverage above a configured limit, relax the $15 minimum, or substitute a different exit plan to make it fit. After any fill, including a partial fill, reconcile the actual filled quantity and orders; only verified fills establish exposure. ## MULTI-TARGET EXITS For every new position that fills, use the actual weighted-average opening fill price as the sole reference for all target and initial-stop calculations. Bind exit triggers to the filled position and applicable quantity. Do not calculate levels from a pre-order mark or estimate when actual fill data is available. Set three profit targets at successive, distinct, favorable EMA/support-resistance levels identified from completed 1-hour/4-hour candle structure and the relevant EMAs. For a long, use resistance levels above entry; for a short, use support levels below entry. Select the next relevant levels in order of distance from entry and keep the selected target prices fixed after entry. If three distinct favorable target levels cannot be identified from available data, do not open the position. At TP1, close 30% of the original filled position quantity. At TP2, close 40% of the original filled quantity. At TP3, close 20% of the original filled quantity. Leave the final 10% as a runner. Reverse the price direction for short positions. Reconcile actual fills and remaining quantity at every exit; never request more than the remaining position. Create the TP1, TP2, and TP3 position-bound price-threshold exit triggers for these quantities after confirmed entry. Keep the existing initial protective stop 3% from the actual weighted-average entry price until TP1 executes. When TP1 is confirmed filled, move the protective stop for all remaining quantity to breakeven at the actual weighted-average entry price. After that, trail the stop using completed 1-hour/4-hour EMA and candle-structure support/resistance levels: for a long, ratchet it beneath relevant rising support; for a short, ratchet it above relevant falling resistance. Only move the stop in the more protective direction. Do not loosen it, and do not move it past the current market in a way that would make it invalid. If no defensible updated level is available, keep the existing protective stop unchanged. The runner has no separate fixed-price take-profit; manage it with the same ratcheting protective stop. If an exit is partial or fails, reconcile fills and position size, handle any remaining required close, and verify the resulting position before considering new entries. After TP1, keep the breakeven/trailing stop active for all remaining quantity while TP2 and TP3 remain active for their specified quantities. After TP3, continue managing the 10% runner with the protective stop. ## EXIT LIFECYCLE AND TIME STOP On every successful evaluation while a position is open, before considering new entries, reconcile actual position side and size and verify that all applicable position-bound stop and target triggers are active, at correct levels, and will not expire before the next scheduled evaluation. Renew or recreate any missing, mismatched, or near-expiry trigger before other actions. Never finish a successful evaluation with an open position lacking its applicable protective stop and remaining target triggers. When an exit trigger fires, treat its requested action as pending and prioritize it on the first successful evaluation. First reconcile current position and completed fills to avoid duplicating a completed close; execute any still-unfilled required close at market and verify the resulting position. If an exit action is partial or fails, handle the remaining position before considering new entries. If neither the initial stop nor TP1 has fired within 72 hours of opening, close the remaining position at the next scheduled evaluation after 72 hours. If TP1 has fired, continue with the multi-target and trailing-runner plan instead of applying the time stop. ## HYGIENE AND TRANSITION Do not open and close the same symbol within one hour unless its stop has fired. Do not re-enter a symbol within two hours of a stop-out. On every opening, state leverage, margin, notional, stop distance, and liquidation distance. If no rule is satisfied this turn, do nothing and state whether the setup lacked sufficient overall confluence or a hard execution/risk constraint blocked it. For positions already open when this strategy takes effect, preserve their existing protective-stop levels and do not widen or reset them. Do not retroactively apply the new initial stop or multi-target plan to those positions; continue to honor their existing position-bound exit triggers until they close. Apply this exit plan to positions opened under it. Chat discussion does not place trades, force an evaluation, or activate this strategy. Saved settings proposals require user approval before they affect autonomous turns.