菜单
First-read routing skill for this quant-skills repository. Use when onboarding into the repo, deciding which quant skill to invoke, installing local skills, or finding the correct path from project instructions to starter-pack to BNB/CMC/alpha research workflows.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | quant-skills-instructions |
| description | First-read routing skill for this quant-skills repository. Use when onboarding into the repo, deciding which quant skill to invoke, installing local skills, or finding the correct path from project instructions to starter-pack to BNB/CMC/alpha research workflows. |
Use this skill first when it is installed.
It routes the agent into the right local instruction layer without loading every quant reference file.
instructions.md for repository layout, source policy, and research standard.start-here/ for readiness checks, provider setup, and lane routing.Need data/API setup:
start-here/
Need Binance raw and clean data files:
alpha-gen-skills/data-download-clean/
Need BNB-specific alpha with anomaly heatmaps:
alpha-gen-skills/bnb-alpha-research/
Need generic OHLCV event-study research:
alpha-gen-skills/ohlcv-alpha-research/
Need CMC/BNB Track 2 strategy-spec packaging:
cmc-bnb-skills/bnb-project-foundation/
Need CMC current market context:
cmc-bnb-skills/cmc-mcp/
Do not create more role files. Prefer scripts for deterministic data pulls, feature engineering, event studies, heatmaps, and candidate JSON.
Run adversarial, leakage-safe quantitative alpha research on OHLCV or related market data. Use when creating a research cycle, auditing price data, defining causal features and targets, registering hypotheses, testing predictive information, validating a trading candidate chronologically, modeling fees and execution, preserving failed experiments, or deciding whether an effect is rejected, informative, paper-tradeable, or a candidate executable alpha.
Pull real Binance BNB market data and run BNB-specific quantitative anomaly research with event studies, regime splits, heatmaps, and backtestable strategy-candidate JSON. Use when asked to derive BNB alpha, analyze BNBUSDT, produce anomaly heatmaps, test BNB breakout/mean-reversion/volatility patterns, or prepare a CMC/BNB Track 2 strategy spec from real free market data.
Use when downloading and cleaning zero-dollar market, derivatives, metadata, and context datasets for alpha generation or backtesting in this quant-skills repo. Covers Binance public klines, futures funding, open interest, long/short ratios, exchange metadata, and optional CoinMarketCap context data when credentials are available.
Build, audit, and interpret instrument-agnostic OHLCV alpha research workflows using only timestamp, open, high, low, close, and volume data. Use for anomaly discovery, past-only feature engineering, event studies, forward targets, MFE/MAE, triple-barrier logic, regime splits, threshold robustness, walk-forward validation, cost stress, and candidate alpha dossiers. Trigger when users ask to research OHLCV alpha, test candle/volume/range anomalies, validate a trading rule, or convert raw OHLCV bars into a disciplined quant research report.
CMC/BNB Track 2 foundation skill for turning BNB quantitative evidence into a sponsor-aligned, backtestable strategy spec. Use after bnb-alpha-research generates anomaly heatmaps and strategy_candidates.json, or when packaging CoinMarketCap/BNB source policy, provenance, and submission-ready strategy JSON.
Generate deterministic, backtestable crypto trading strategy specs from CoinMarketCap market context and historical OHLCV candles for BNB Hack Track 2 Strategy Skills. Use when asked to create, explain, validate, or demo a CoinMarketCap Strategy Skill, produce strategy JSON, classify market regimes, choose volatility breakout/mean reversion/range strategies, or document data provenance for CMC/free/fixture/synthetic market data.