skills.rest
wudye
wudyeCOMMUNITY
Explore all skills by this author →
GitHub

strategy-generate

Generate and backtest quantitative trading strategies with config.json and SignalEngine code.

Updated May 5, 2026
Search Tags:#performance metrics#market data#signal generation#strategy backtesting#quantitative trading#portfolio allocation#python engine
One-click install
npx skills add https://github.com/wudye/traderAssistHK --skill strategy-generate-wudye
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: strategy-generate
Source: https://github.com/wudye/traderAssistHK/tree/main/backend/src/skills/strategy-generate
Command: npx skills add https://github.com/wudye/traderAssistHK --skill strategy-generate-wudye

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill turns vague trading ideas into a runnable quantitative strategy by generating the required configuration and strategy signal code, then guiding users through backtest-driven iteration so they can improve results.

Core Features & Use Cases

  • Requirements parsing to configuration: converts user intent into a config.json specifying instruments, dates, and execution/backtest settings.
  • Signal engine code generation: produces a code/signal_engine.py implementation that follows the SignalEngine contract and outputs position signals in the required range.
  • Backtest evaluation workflow: runs the engine’s built-in backtest process and evaluates artifacts like artifacts/metrics.csv and artifacts/equity.csv against clear gates (e.g., metrics existence, non-empty equity, non-NaN equity, and trade count).
  • Multi-market support: supports China A-shares, US stocks, Hong Kong stocks, and cryptocurrencies with consistent code normalization rules and an auto-routing source strategy.

Quick Start

Use the strategy-generate skill to build a 5-day vs 20-day dual moving-average crossover strategy for 000001.SZ, and backtest it for 2024.

Frequently Asked Questions about strategy-generate

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate and backtest a quantitative trading strategy from a natural language idea?▼

To generate and backtest a quantitative trading strategy, this Skill converts your requirements into a config.json and SignalEngine-compatible pandas/numpy signal code, then runs an iterative backtest to produce performance metrics and equity artifacts.

Can I backtest trading strategies for both cryptocurrencies and equities using the same engine?▼

Yes, you can backtest trading strategies across China A-shares, US stocks, Hong Kong stocks, and cryptocurrencies, as the engine applies consistent code normalization and an auto-routing source strategy for multi-market support.

How does signal generation handle position sizing and index alignment for portfolio allocation?▼

Signal generation requires deterministic pure pandas/numpy implementations to output position signals with strict index alignment, ensuring accurate entry/exit logic and portfolio allocation within the generated SignalEngine contract.

What are the evaluation gates for validating backtest artifacts like metrics and equity curves?▼

Backtest evaluation validates generated artifacts by checking for metrics.csv and equity.csv existence, non-empty and non-NaN equity values, and confirming at least one trade was executed to ensure non-empty performance metrics.

What is the best way to iterate on a dual moving-average crossover strategy after an initial backtest?▼

The best way to iterate on a dual moving-average crossover strategy is through artifact-based assessment, modifying the generated signal code and config.json based on evaluated performance metrics and equity artifacts to improve backtest results.

Related Skills

Explore More Finance & Accounting Skills→
MoyuchiiiiMoyuchiiii

backtest-runner

Run backtests on specified stocks and time periods to output returns, drawdown, and Sharpe ratio.

Community
Intermediate
CambixxCambixx

trading-plan-generator

Generates structured trading plans with risk management, sizing, and tracking.

Community
Intermediate
20YN0420YN04

candlestick

Detects 15 classic candlestick patterns and generates composite trading signals from OHLCV data.

Community
Intermediate
0xZKnw0xZKnw

strategy-generate

Automate design, coding, and backtesting of quantitative trading strategies.

Community
Advanced