メニュー
交易系统开发知识库，基于《Trading Systems》(Tomasini & Jaekle 2009)。支持：(1) 趋势跟踪/均值回归策略设计；(2) Walk Forward Analysis 参数优化；(3) 仓位管理（固定分数、固定比率）；(4) 组合构建（权益线交叉、相关性管理）。触发于：开发交易系统、设计量化策略、参数优化验证、回测评估、仓位管理方法、多策略组合。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | trading-systems |
| description | 交易系统开发知识库，基于《Trading Systems》(Tomasini & Jaekle 2009)。支持：(1) 趋势跟踪/均值回归策略设计；(2) Walk Forward Analysis 参数优化；(3) 仓位管理（固定分数、固定比率）；(4) 组合构建（权益线交叉、相关性管理）。触发于：开发交易系统、设计量化策略、参数优化验证、回测评估、仓位管理方法、多策略组合。 |
基于《Trading Systems: A new approach to system development and portfolio optimisation》的知识库。
查询策略类型：
查看评估标准：
想法 → 编程 → 回测 → 评估 → 优化 → Walk Forward → 仓位管理 → 组合构建
详见 system_development.md 完整流程。
| 指标 | 优秀标准 |
|---|---|
| Profit Factor | > 1.5 |
| Average Trade | > 滑点 + 佣金 |
| RINA Index | > 30 |
| Max Drawdown | 尽可能低 |
| 文件 | 内容 |
|---|---|
| trend_following.md | 通道突破、均线交叉、LUXOR 案例 |
| mean_reversion.md | 布林带、三角形系统 |
| position_sizing.md | 固定分数、固定比率、Monte Carlo |
| risk_management.md | 止损、跟踪止损、评估指标 |
| system_development.md | 设计→测试→优化→Walk Forward |
| portfolio_construction.md | 分散化、权益线交叉法 |
examples/python/trend_following.py - 趋势跟踪系统examples/python/position_sizing.py - 仓位管理方法examples/python/walk_forward.py - Walk Forward AnalysisFull_doc/full_content.md - 全文（含 Figure 描述）Full_doc/figures/ - 135 张原图Full_doc/figures_metadata.json - Figure 元数据索引index.json - 35+ 词条索引glossary.md - 250+ 条术语表（ATR、WFA、Kelly公式等）概念查询：先查 index.json 定位文件，再读对应 reference。
代码生成：参考 examples/python/ 目录下的示例。
流程指导：读 system_development.md 获取步骤清单。
使用 akshare 获取中国金融市场实时数据和历史数据。当需要查询 A 股、港股、美股、指数、基金、期货等金融产品的实时行情、历史数据、财务报表时使用该技能。
Senior high-frequency factor construction workflow for turning a market mechanism into a causally valid, measurable, normalized, and implementable factor design. Trigger when the user asks to construct, formalize, document, or pseudo-implement microstructure / order flow / orderbook / event-time alpha factors, maker toxic-flow filters, liquidation cascade factors, OFI / imbalance signals, Hawkes smoothing design, or binary-mask factor decomposition. Do NOT use for pure backtest optimization, factor mining/search, label design, model selection, feature importance, or full strategy validation. Core output is a fixed research contract: mechanism, data contract, temporal contract, dimensions, mask semantics, aggregation, normalization, smoothing, observability risk, must-fail scenarios, and minimal sanity checks.
Industrial-grade workflow for developing, testing, and verifying financial ML strategies based on "Advances in Financial Machine Learning" (AFML) and "Machine Learning for Asset Managers" (MLAM) by Marcos López de Prado. Use when: (1) Building Dollar/Volume/Imbalance bars, (2) Triple-barrier or meta-labeling, (3) Sample weights and uniqueness, (4) Purged/embargoed cross-validation, (5) Feature importance (MDI/MDA/Clustered MDA), (6) Trend scanning, (7) Backtest verification (DSR/PSR/CPCV), (8) HRP portfolios, (9) Fractional differentiation, (10) CUSUM filtering, (11) Market microstructure analysis. Includes causal verification framework with validation metrics and book references. For code-level implementation, see afmlkit skill.
Evaluate whether a design pattern is warranted during feature design, refactoring, architecture changes, or code review, then choose a fitting pattern and a Python-idiomatic implementation. Use when code shows change pressure such as repeated if/elif branching, tangled object creation, oversized classes, mixed responsibilities, unstable dependencies, event fan-out, or requests to "apply a design pattern", "improve extensibility", "decouple modules", or "make this easier to evolve".
Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of "Word doc", "word document", ".docx", or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a "report", "memo", "letter", "template", or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.
A specialized Financial Machine Learning knowledge expert based on the works of Marcos López de Prado. Use when the user asks for theoretical explanations, methodology, or best practices in Financial ML (e.g., Triple-Barrier Method, Fractional Differentiation, Hierarchical Risk Parity).