AI Agent Hub
Back to skills
Dong Formula Editor icon

Dong Formula Editor

Professional Updated 2026.08.30

Paste the following prompt into your AI chat to install this skill:

Please install @user_4f3fa403/dong-formula according to https://skillhub.cn/install/skillhub.md.

About this skill

Problem

Writing trading indicators often fails because of platform-specific syntax, not because the idea is hard. In Dong formulas, small details matter: template placeholders may be left behind, EMA uses a non-standard smoothing rule, KDJ should use SMA(RSV,3,1), MACD() returns only DIF, division by zero can return NaN, and future-looking functions can introduce look-ahead bias. This skill treats indicator writing as a structured task rather than ad hoc script assembly.

How It Works And Limits

The skill routes requests into three modes:

  • Template metrics: When the user mentions known indicators such as MACD, KDJ, or Bollinger Bands, it matches the template, extracts parameters, replaces {KEY} placeholders, and outputs the formula with parameter notes.
  • Custom formulas: When the user describes intent, such as marking three consecutive up days or plotting yesterday's high, it selects the appropriate functions and drawing commands from the Dong language reference.
  • Syntax Q&A: When the user asks about CROSS, arrow drawing, or function signatures, it returns the signature, explanation, and a runnable example.

Key constraints to watch: warm-up periods may return NaN instead of zero; colors use uppercase hexadecimal values without a leading #; denominators should avoid division by zero; and functions such as BACKSET, ZIGZAG, and PEAK are future functions, so their output should be clearly flagged.

Use Cases

  • Replace template indicator parameters for MACD/KDJ before backtesting.
  • Generate a custom draw-arrow formula when price rises for three consecutive days.
  • Look up CROSS/EMA signatures and get runnable examples while debugging.
  • Check division-by-zero, NaN, and future-function rules before publishing.

Best For

  • Researchers maintaining Dong indicator strategies who need template indicators parameterized for backtesting.
  • Engineers writing custom charting logic who need to map market intent to functions and draw commands.
  • Quant developers debugging missing plots or parameter errors who need signatures, examples, and platform rules.
  • Risk reviewers checking published indicators for future functions, NaN handling, and division-by-zero guards.