noahnan-max/chanlun-trading-system
Developer toolsThis page was translated automatically. Technical details may differ from the original. View the original
Executable and verifiable Chan Theory (Chan Zhong Shuo Chan) AI Agent Skill — Determine the level first, identify the structure first, and first write the failure points. For research/study purposes, not recommended stocks. An operational Chan Theory (Chan Theory) Skill for AI agents; research only.
Overview
Make **'s technical analysis system a system that can be directly integrated into an AI Agent—let AI help you read the movements of A-shares, Hong Kong stocks, ETFs, indices, and futures, instead of just shouting "must-buy points." The core is a ** pattern: Level gate → Structure gate → Type gate → Comparison gate → Buy and sell point gate → Trigger gate → Risk gate → Downgrade gate. If any step is missing, the output automatically returns from "Confirm Buy" back to "Observe." The above remote command assumes that GitHub v0.1.1 tag has already been released; Before release, please use the "Local Development and Installation" option below. Public market data is an optional convenient entry point; CSV and built-in examples do not rely on it: Codex/OpenAI users can have $skill-installer install chanlun-trading-system from this GitHub repository, or download chanlun-trading-system-skill.zip from GitHub Release and then execute: Tencent SkillHub uses chanlun-trading-system-skillhub.zip from the same release; it only adds the distribution metadata required by SkillHub than the general package, with the rule body and scripts aligned. Skill and the visual workbench are two installation units at runtime: only Skill can be installed for standardized text research; To open the interaction chart, you also need to install chanlun-visual. Skill will first run doctor; if missing, it only prompts installation and will not silently modify the environment. You can extract the same release version of Skill into ~/.claude/skills/. Local cloning installation method: Other agents or local models can also load SKILL.md as system prompts, and when details are needed, then read references/*.md by table.
Overview (machine translated)
Turn the technical analysis system of Chan Zhongshuo Chan (Chan Theory) into a Skill that can be directly integrated into an AI Agent—let AI help you read the movements of A-shares, Hong Kong stocks, ETFs, indices, and futures. First, first set the level, identify the structure, and first write the failure points, rather than just shouting “guaranteed buy points.”
⚠️ > ** For technical research and study only, does not constitute investment advice, does not provide buy or sell orders, and does not promise returns. ** Chan Theory is subject to subjectivity and invalidation risks; profits and losses are borne by the user.
What exactly does this skill do?
| Not | Yes |
|---|---|
| ❌ A “Sure Rise Signal” Machine | ✅ A Mandatory Discipline Analysis Process |
| ❌ Indicator Golden Cross Calls for Buy | ✅ After completing the → fractal→ → line segments→ the central → divergence→ only draw conclusions |
| ❌ Treat the breakout as a third buy | ✅ Central zone exit + pullback + no pullback—only when all are named is it recognized as a third buy |
| ❌ Pretend to understand any chart | ✅ If the data is insufficient, downgrade to “Observe” and explain the approximate loss |
The core is a Audit Gate: Level Gate → Structure Gate → Type Gate → Comparison Gate → Buy and Sell Point Gate → Trigger Gate → Risk Gate → Downgrade Gate. If any step is missing, the output automatically returns from “Confirm Buy” back to “Observe.”
Installation
1. Direct operation of the visual workbench
After installing uv, you can run it all at once from a fixed version without degrading your existing Python environment:
Long-term use is required:
The above remote commands assume the GitHub v0.1.1 tag has been published; Before publishing, please use the “Local Development and Installation” option below. Public market data is an optional convenient entry point; CSV and built-in examples do not depend on it:
2. Install AI Skills
Codex/OpenAI users can have $skill-installer install chanlun-trading-system from this GitHub repository, or download chanlun-trading-system-skill.zip from GitHub Release and then run:
Tencent SkillHub uses chanlun-trading-system-skillhub.zip from the same release; It only adds the distribution metadata required by SkillHub compared to the general package, and the rule body and script are consistent.
Skill and the visual workbench run as two installation units: only Skill can be installed for standardized text research; To open the interaction chart, you also need to install the above chanlun-visual. Skill will first run doctor, and if missing, only prompts installation and will not silently modify the environment.
Claude Code can extract the same release Skill to ~/.claude/skills/. Local cloning installation method:
Other agents / local models can also load SKILL.md as a system prompt, and read references/*.md by table when details are needed.
This page was translated automatically. Technical details may differ from the original.
Recommended Tools
Try a different keyword or remove a filter.
Install
npx skillfish add noahnan-max/chanlun-trading-system