🐺 DIRE v1.0 — BRENTOIL XYZ Specialist

One asset. News-driven. Scanner enters. DSL exits.

---

⛔ CRITICAL AGENT RULES

RULE 1: Install path is /data/workspace/skills/dire-strategy/

RULE 2: THE SCANNER DOES NOT EXIT POSITIONS

When the scanner sees an active BRENTOIL position, it outputs NO_REPLY. DSL is the ONLY exit mechanism. Kodiak-family lesson: scanner-driven thesis exits fight DSL tier locks and chop positions before they can mature. DSL owns all exits.

RULE 3: MAX 1 POSITION — xyz:BRENTOIL only

No multi-asset rotation. No cross-asset hedges. Dire trades oil, period.

RULE 4: XYZ asset — coin field MUST be xyz:BRENTOIL

The xyz: prefix is mandatory for Hyperliquid HIP-3 DEX. Passing plain BRENTOIL will fail at the exchange. The scanner auto-applies the prefix; verify.

RULE 5: ISOLATED margin only

XYZ DEX does not support CROSS margin. Scanner forces leverageType: "ISOLATED". Do not attempt to change this.

RULE 6: Scanner output is AUTHORITATIVE

The JSON emitted to stdout is the source of truth for what happened on a scan. Runtime consumes it; logs trust it.

RULE 7: Verify runtime is installed on every session start

Run openclaw senpi runtime list. Runtime must be listed. The position tracker and DSL exit engine are handled by the plugin runtime via runtime.yaml.

RULE 8: Leverage scales with conviction. Hard-capped at 10x. One position. Period.

v1.1: Sizing is tiered by score, not fixed. Higher conviction → more leverage AND more margin, enabling big winners to pay for fees and small losers:

Score	Leverage	Margin	Notional / account	Tier
9	3x	20%	0.6x	cautious (just cleared MIN)
10	5x	25%	1.25x	standard (fleet baseline)
11	7x	30%	2.1x	conviction
12+	10x	30%	3.0x	apex

10x cap = 50% of Hyperliquid's 20x BRENTOIL max. ISOLATED margin means each trade's risk is bounded by its own margin allocation. DSL triggers in minutes at 10x (a 0.5% price move = +5% ROE → T0 lock).

RULE 9: Drawdown circuit breaker at 15% from 7-day rolling peak

Scanner auto-pauses HUNTING if account_value / peak_7d < 0.85. Auto-resumes when account recovers to >90% of peak, OR operator sets dd_manual_unlock: true in state/trade-counter.json.

RULE 10: Wolverine execution pattern

Scanner calls MCP via mcporter CLI subprocess directly. No LLM parse loop for entry execution. Required for news-latency-sensitive oil entries. DSL attach happens inline before scanner returns — if DSL attachment fails, position is immediately closed to prevent unprotected exposure.

---

Thesis

Oil responds to discrete news events (OPEC decisions, geopolitical conflict, inventory releases) with sharp directional moves. Typical clean move: 2-5% over 1-4 hours. Typical reversal after overshoot: 3-8% correction over 30 minutes to 2 hours.

Dire's edge: catch the clean directional move early (4TF alignment + SM consensus

• volume spike + OI acceleration), ride with DSL tier protection, exit on SM flip or DSL-locked profit floor.

What Dire does NOT do:

• Predict news before it happens
• Fade overextended moves (that's Vulture's thesis on crypto)
• Trade multiple XYZ assets (one asset, one thesis)
• Pyramid into winners (one position max)

Architecture

Single-asset specialist, 3-mode state machine (HUNTING/RIDING/STALKING), 4TF alignment hard gate, SM HARD BLOCK, OI velocity scoring, Wolverine execution pattern, tiered DSL. Ported from Kodiak v5.0 with XYZ-specific adaptations.

State machine

Mode	Trigger	Behavior
HUNTING	No open position	Scan every 60s; fire create_position via Wolverine pattern when all gates pass
RIDING	Open position, ROE < Tier 0 (+5%)	Let position run; DSL engine not yet active
STALKING	Open position, ROE ≥ Tier 0	DSL locked via ratchet_stop_add; scanner watches for SM flip exit

Scan cadence

60 seconds (cron * * * * *). Faster than crypto's 3-min cadence because oil news breaks faster than crypto on-chain signals.

DSL ladder (5 tiers, aggressive early lock)

Tier	Trigger ROE	Lock ROE (% of HW)
T0	+5%	25%
T1	+10%	50%
T2	+20%	70%
T3	+35%	80%
T4	+50%	90%

Aggressive early lock at T0 (+5% → 25% HW) protects against geopolitical V-reversals like the Apr 17 BRENTOIL wick ($108 → $89 → $97 in 4 hours).

Sizing (v1.1 — conviction-scaled)

Fixed caps:

Parameter	Value
absolute max leverage	10x (50% of HL's 20x BRENTOIL max)
leverageType	ISOLATED (required for XYZ)
positions max	1

Tier table (leverage AND margin scale with score):

Score	Leverage	Margin %	Notional / account	Tier label
9	3x	20%	0.6x	cautious
10	5x	25%	1.25x	standard
11	7x	30%	2.1x	conviction
12+	10x	30%	3.0x	apex

Gate order (all must pass)

1. Drawdown circuit breaker (account ≥ 85% of 7-day peak)
2. Daily entry cap (max 2 per 24h)
3. 4TF alignment HARD GATE (5m, 15m, 1h, 4h all same direction)
4. SM HARD BLOCK (SM direction must match)
5. SM conviction scoring (v1.1): premium |>0.1%| → +1, |>0.3%| → +2
6. OI velocity: oi_velocity.oi_change_pct_1h > +5 → +2, > +2 → +1, < -3 → -1, null passes
7. Volume ratio (v1.1 tiered): 15m vol / 1h avg > 2.5x → +1, > 5x → +2 (extreme news)
8. Price action cleanliness: no 5m wicks > 1.5% against direction in last 30 min → +1
9. MIN_SCORE ≥ 9 (tunable)

Max attainable score: 6 (base for 4TF+SM aligned) + 2 (SM conviction) + 2 (OI) + 2 (vol) + 1 (clean) = 13. Apex tier (score 12+) is rare by design.

Execution

Wolverine pattern: Python scanner invokes mcporter CLI directly for create_position and ratchet_stop_add. Bypasses LLM parse loop. Validates inner-order JSON in-process.

Entry order config:

• orderType: "FEE_OPTIMIZED_LIMIT"
• feeOptimizedLimitOptions: { ensureExecutionAsTaker: true, executionTimeoutSeconds: 10 }
• No slippagePercent (incompatible with FOL)

DSL attach is mandatory and inline — if ratchet_stop_add fails after create_position succeeds, scanner immediately closes the position via close_position (orderType MARKET, 2% slippage) to prevent unprotected exposure.

Exits (priority order)

1. DSL tiered ladder (primary, 90%+ of exits)
2. SM flow flip (secondary) — market close immediately
3. 4TF break — tertiary, if both 1h and 4h flip against position
4. No hard TP — DSL replaces TP entirely

Configuration

• config/dire-config.json — wallet, strategy ID, MIN_SCORE, starting budget
• runtime.yaml — DSL tier ladder (must match scanner's ladder), position tracker
• scripts/dire_config.py — constants, MCP helpers, state I/O
• scripts/dire-scanner.py — main scanner loop

State files

Written to state/ inside the skill directory:

• trade-counter.json — daily entry count, drawdown gate state, last results
• state.json — mode, last scan ts, 7-day rolling peak

Monitoring

72-hour checkpoint

Green: ≥1 entry attempted, DSL attached cleanly, no ghost trades, equity ≥ 95%

Yellow: Scanner firing every 60s but no valid setups (gates too tight OR current market not producing oil momentum)

Red: Ghost trade, DSL attach failure, account drops >15% (circuit breaker should have triggered)

Week 1 review

• Trade count: 3-10 entries expected (2/day cap × 7 days × ~40% gate hit rate)
• Win rate: ≥50% at >2:1 avg_win:avg_loss ratio
• Fees as % of gross PnL: <25%

Version history

• v1.1 — Conviction-scaled sizing. Leverage 3x→10x and margin 20%→30% scale with score. Added SM conviction scoring (premium tier bonus) and volume spike tier-2 (>5x for extreme news). Fixes the v1.0 failure mode where capped winners
  • tight DSL + normal fee drag would produce slow bleed even on 50%+ win rates. Apex setups (score 12+) now deploy up to 3x account notional.
• v1.0 — Initial release. First XYZ specialist. Kodiak-family port. Fixed 3x leverage / 30% margin. Deprecated — capped upside couldn't pay for fees.