Agent SkillsSuperior-Trade/superior-skills › related-market-spread

related-market-spread

GitHub

用于检测 Polymarket 关联市场间的定价差异,生成基于价差阈值的相对价值交易策略原型。支持通过历史成交数据回溯测试价差行为,并指导用户将原型适配为可部署的 NautilusTrader 代码,同时提示流动性与执行风险。

skills/v3/polymarket-archetypes/related-market-spread/SKILL.md Superior-Trade/superior-skills

Trigger Scenarios

用户询问两个关联市场的相对价值或定价偏差 请求检查 linked outcomes 的价格不一致性 寻找 mispriced 市场相对于另一市场的套利机会

Install

npx skills add Superior-Trade/superior-skills --skill related-market-spread -g -y
More Options

Non-standard path

npx skills add https://github.com/Superior-Trade/superior-skills/tree/main/skills/v3/polymarket-archetypes/related-market-spread -g -y

Use without installing

npx skills use Superior-Trade/superior-skills@related-market-spread

指定 Agent (Claude Code)

npx skills add Superior-Trade/superior-skills --skill related-market-spread -a claude-code -g -y

安装 repo 全部 skill

npx skills add Superior-Trade/superior-skills --all -g -y

预览 repo 内 skill

npx skills add Superior-Trade/superior-skills --list

SKILL.md

Frontmatter
{
    "name": "related-market-spread",
    "updated": 1781654400,
    "version": "0.1.0",
    "description": "Use when two Polymarket markets imply different probabilities for a linked outcome."
}

Strategy: Polymarket · Related-Market Spread

When to use

Use this when the user asks for relative-value trades, discrepancy checks, linked outcomes, or questions like "this market is mispriced versus that one."

What the agent should look for

  • At least two exact market slugs from POST /v3/markets/search.
  • Resolution logic that is genuinely comparable across the two markets.
  • Spread between filled-trade prices that is wide versus historical behavior.
  • Enough historical filled data on both markets for the requested backtest window.
  • Ability to enter and exit both legs (or acknowledge unavailable liquidities before entering).

Backtest fit with filled data

Partial fit. Backtests built from filled TradeTick data are useful for checking whether the spread relationship has widened and then compressed in history. They cannot guarantee simultaneous fills or full cross-book liquidity.

Use this as a generated archetype: the strategy should be adapted in code before use, not treated as a finished edge.

Strategy logic

Track the latest filled price for both legs. Enter when spread is above a threshold and exits when spread narrows or max holding duration expires.

Nautilus strategy shape

  • Subscribe to TradeTick for both instruments.
  • Keep the latest filled price per instrument.
  • Compute spread each tick using the latest values.
  • Enter and flatten only when spread thresholds are hit.
  • Manage stale-state safety (e.g., skip entries until both legs have recent fills).

Example strategyConfig

{
  "primary_instrument_id": "AUTO_FROM_MARKET_A",
  "secondary_instrument_id": "AUTO_FROM_MARKET_B",
  "spread_entry": 0.08,
  "spread_exit": 0.03,
  "order_size": 10,
  "max_holding_ticks": 120
}

Iteration knobs

Knob Effect
spread_entry Higher = only trade stronger relative-value discrepancies.
spread_exit Lower = wait for fuller normalization before exiting.
max_holding_ticks Lower = tighter time risk control if convergence does not happen.

Failure modes

  1. Markets with different outcome mappings can produce a real structural spread, not an inefficiency.
  2. Resolution rules can shift, changing what "discrepancy" means.
  3. One leg may look tradable in backtest and be unavailable or illiquid live.
  4. TradeTick replay cannot prove simultaneous execution across legs.

User-facing framing

"This is a relative-value archetype, not a guaranteed edge. It is built from filled TradeTick backtests and exact market slug discovery first, then adapted into custom NautilusTrader code before any deployment."

Version History

  • 85f77be Current 2026-08-02 21:45

Same Skill Collection

skills/external-deposit/SKILL.md
skills/v2/exchanges/aerodrome/SKILL.md
skills/v2/exchanges/hyperliquid/SKILL.md
skills/v2/primitives/backtesting/SKILL.md
skills/v2/primitives/fees-optimizations/SKILL.md
skills/v2/primitives/trade-thesis/SKILL.md
skills/v2/strategies/basis-arb/SKILL.md
skills/v2/strategies/bollinger-reverter-4h/SKILL.md
skills/v2/strategies/breakout/SKILL.md
skills/v2/strategies/dca-weekly/SKILL.md
skills/v2/strategies/donchian-strong-regime/SKILL.md
skills/v2/strategies/funding-rate-arbitrage/SKILL.md
skills/v2/strategies/funding-squeeze/SKILL.md
skills/v2/strategies/grid-trading/SKILL.md
skills/v2/strategies/mean-reversion/SKILL.md
skills/v2/strategies/scalping/SKILL.md
skills/v3/exchanges/lighter-robinhood/SKILL.md
skills/v3/exchanges/lighter/SKILL.md
skills/v3/exchanges/polymarket/SKILL.md
skills/v3/polymarket-archetypes/catalyst-confirmation/SKILL.md
skills/v3/polymarket-archetypes/deadline-drift/SKILL.md
skills/v3/polymarket-archetypes/large-fill-pressure/SKILL.md
skills/v3/polymarket-archetypes/probability-mean-reversion/SKILL.md
skills/v3/polymarket-archetypes/probability-momentum/SKILL.md
skills/v3/primitives/deposit-qr/SKILL.md
skills/v2/primitives/dsl-exit-engine/SKILL.md
skills/v2/primitives/intelligence/SKILL.md
skills/v2/primitives/regime-overlay/SKILL.md

Metadata

Files
0
Version
201271f
Hash
f2174faa
Indexed
2026-08-02 21:45

ホーム - Wiki
Copyright © 2011-2026 iteam. Current version is 2.155.2. UTC+08:00, 2026-08-12 23:20
浙ICP备14020137号-1 $お客様$