SuperAPI AI Directory

mongoose-strategy

General & other

View repository

Updated 10-16 08:46

About this skill

🪙 MONGOOSE v1.0 — On-Chain Finance vs Legacy

Mongoose bets that money is migrating on-chain — stablecoins, crypto exchanges, and BTC treasuries are eating legacy financial rails. It expresses that as a long/short:

  • LONG the on-chain rails — HYPE (the venue), CRCL (Circle/stablecoins), COIN (Coinbase), HOOD (Robinhood), MSTR + PURRDAT (crypto treasuries).
  • SHORT legacy finance — BX (Blackstone) + SP500 (broad financial-beta).

The P&L is the disruptor-vs-incumbent spread. One producer (mongoose-producer.py) serves both books; MONGOOSE_LEG selects.

Book Holds Wallet env Runtime Scanner
long on-chain rails (LONG) MONGOOSE_LONG_WALLET runtime-long.yaml mongoose_long_signals
short BX + SP500 (SHORT) MONGOOSE_SHORT_WALLET runtime-short.yaml mongoose_short_signals

The thesis, asset by asset

Sleeve Direction Names Rationale
The venue LONG HYPE (1.3×) the on-chain exchange itself — core of the thesis
Stablecoin rails LONG CRCL (1.2×) Circle/USDC — the clearest on-chain-finance winner
Crypto brokers/exchanges LONG COIN, HOOD Coinbase, Robinhood
Crypto treasuries LONG MSTR (0.7×), PURRDAT (0.6×) BTC/HYPE treasury proxies — sized down (levered, double-count)
Legacy finance SHORT BX (0.8×), SP500 (1.2×) the incumbents + broad financial-beta — the hedge

The honest caveat — the hedge is loose

The short leg is cross-sector and thin: few pure legacy-finance names (banks, payment incumbents) list on the venue yet, so the hedge leans on SP500 + the one liquid legacy name (BX). It nets down market beta rather than perfectly isolating disruptor-vs-incumbent. As more legacy-finance names list, add them to config.universe on the short book to sharpen the pair. The long book is the strong, high-conviction side.

How it picks (producer-side)

  1. Curated universe intersected with the live board + a relative-to-market liquidity gate (24h vol ≥ volFloorPctOfMedian × the universe median — no $ floor).
  2. ABSOLUTE trend is the gate — long a rail only while its 4h is not bearish; short an incumbent only while its 4h is not bullish (+ capitulation guard).
  3. Relative strength is a tiebreaker — tilts ranking/size, never benches a trender.
  4. Conviction sizing — margin = account_value × marginPct × sizingWeights[name]. No $.

The long/short balance is YOUR dial

The split is the operator's funding across the two wallets (config _hedge_note) — not hardcoded. Default a net-long tilt (the thesis is bullish the disruptors; long book 5 slots / 18% / 5x, legacy book 2 slots / 15% / 4x). Fund 50/50 for a tighter (still cross-sector) hedge.

Fleet-standard rules (enforced)

  • Max leverage 5x on-chain / 4x legacy; per-position margin ≤ 18% / 15% (× weight).
  • Drawdown halt 20% / 18%, reset_on_day_rollover.
  • Mandatory DSL; entries + exits FEE_OPTIMIZED_LIMIT with taker fallback.
  • Verbose per-tick JSON; sizes off max(main, xyz) account value.
  • Legacy book runs tighter (lower leverage, tighter max-loss, faster stall-cuts).

Hard rule — user-conversation sessions are READ-ONLY

A Claude session conversing with a user MUST NOT call create_position, close_position, edit_position, ratchet_stop_*, cancel_order, or any strategy_close* tool against Mongoose's wallets. Entries are emitted only by the producer daemon; exits are owned only by the runtime DSL.

v1.0 — initial build

A Lion-family two-book long/short pointed at the on-chain-finance disruption. Strong long book (all hot, live names); the short is the deliberately-thin, index-led hedge — documented honestly — to be sharpened as more legacy-finance names list on the venue.