Skip to content
Get packs

Guide · Agent skill

How to use the Candlestick Pattern Recognition agent skill

Updated · written with AI from the project's own documentation and checked against our catalogue data

The candlestick skill gives an AI agent a rules-based library of 15 classic candlestick chart patterns, plus a simple way to turn those patterns into a long, short or stand-aside signal. It is aimed at people learning how candlestick rules are coded, developers prototyping a trading strategy, and analysts who want a quick scan of historical OHLC price data. By the end of this guide you will know what the skill does, how to install it, how to ask for it in plain language, and what to check before trusting its output.

Candlestick Pattern RecognitionHKUDS/Vibe-Trading · MIT

Quick start

  1. 1Download the ZIP. Download the candlestick skill ZIP from this page.
  2. 2Unzip into a skills folder. Unzip it into ~/.claude/skills/ for all projects, or into a project's .claude/skills/ folder for one project.
  3. 3Or upload in the Claude apps. In the Claude apps, open Settings, go to Capabilities, and upload the skill under Skills.
  4. 4Install the Python dependencies. Run pip install pandas numpy requests in your environment.
  5. 5Prepare OHLC data. Have your open, high, low and close price history ready, with a time index.
  6. 6Ask for a scan in plain language. Describe the patterns or signal you want and point the agent at your data.
  7. 7Check the signal. Review the score and the pattern rules, then backtest before drawing any conclusions.

What is an agent skill and what does this one do?

An agent skill is a folder containing a SKILL.md instruction file and optional supporting files, which an AI agent can load when a task matches the skill's description. This one is named candlestick and is in the strategy category.

The skill teaches the agent to identify 15 candlestick patterns in price data: 5 single-candle patterns, 5 double-candle patterns, 4 triple-candle patterns and 1 trend confirmation. Bullish patterns score +1 and bearish patterns score -1. The total is turned into a signal: 1 for long when the score is above 0, -1 for short when it is below 0, and 0 to stand aside when it equals 0.

  • Single-candle: Hammer, Inverted Hammer, Shooting Star, Doji, Spinning Top
  • Double-candle: Bullish Engulfing, Bearish Engulfing, Bullish Harami, Bearish Harami, Piercing Line, Dark Cloud Cover
  • Triple-candle: Morning Star, Evening Star, Three White Soldiers, Three Black Crows

How do you install the candlestick skill?

Download the ZIP file for the skill from this page. Then either unzip it into your personal skills folder, ~/.claude/skills/, or into a project's .claude/skills/ folder if you want the skill available only in that project.

If you use the Claude apps instead, open Settings, go to Capabilities, and upload the skill under Skills.

The skill's SKILL.md lists its Python dependencies as pandas, numpy and requests. Install them with the command below before running the included script.

pip install pandas numpy requests

What do you need before you start?

You need Python and the libraries listed above. You also need your own OHLC price data (open, high, low, close, and the usual time index). The skill does not come with market data.

  • Python with pandas, numpy and requests installed
  • OHLC price data supplied by you
  • Note: requests is listed in the install command but is not used in the shown script excerpt

How do you trigger the skill with natural requests?

You do not need special syntax. Describe the task in plain language and mention candlesticks, patterns or signals. The agent can match your request to the skill's description and load it.

Name the patterns you care about if you want a narrower answer, and say what data you are using so the agent knows which price series to read.

  • Scan this OHLC CSV for candlestick patterns and tell me the composite signal for each day.
  • Which bullish or bearish candlestick patterns appear in the last 50 bars of this price history?
  • Check whether there is a morning star or evening star in my BTC data and explain the score.
  • Use the candlestick skill to tell me whether the total pattern score is long, short or flat.

What does the example signal engine do?

The included example_signal_engine.py file shows how the pattern scores can be computed quickly across a whole price history using pandas. Read it to see how each rule is coded, then adapt it to your data.

The skill has two adjustable parameters: body_pct, default 0.1, is the threshold for body-to-range ratio used to detect a doji, and shadow_ratio, default 2.0, is the ratio of shadow length to body length.

  • Files included: SKILL.md and example_signal_engine.py
  • body_pct default: 0.1
  • shadow_ratio default: 2.0

What should you check before relying on the output?

The source gives no evidence of profitability and does not validate the pattern rules. The rules are simple and ignore transaction costs, risk management and market context. Treat the output as a starting point for study, not a trading decision.

Backtest any signal on historical data before using it, and be careful about running scripts you have not read. The skill runs scripts, and the source does not describe any network use for this skill.

  • Runs scripts: yes. Read example_signal_engine.py before running it.
  • Uses network: no, according to this skill's description.
  • Check results against your own backtest with realistic costs.

What people use it for

  • Learning how hammers, engulfing bars and morning or evening stars are coded in pandas
  • Prototyping a trading strategy that starts from a candlestick pattern library
  • Running a quick rules-based scan of historical OHLC data for pattern counts and composite signals
  • Comparing how the pattern score changes across different doji or shadow thresholds

Frequently asked questions

Is the candlestick agent skill free to use?

The source does not state a price or licence terms for the skill itself. The repository it comes from is listed under the MIT licence, which you should check directly for the terms that apply.

What do I need to install the candlestick skill?

You need Python with pandas, numpy and requests, installed with pip install pandas numpy requests. You also need your own OHLC price data, since the skill does not include any.

Where do I put the candlestick skill folder?

Unzip the downloaded ZIP into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder for one project. In the Claude apps, you can instead upload it under Settings, Capabilities, Skills.

Which candlestick patterns does this skill detect?

It detects 15 patterns: Hammer, Inverted Hammer, Shooting Star, Doji and Spinning Top; Bullish and Bearish Engulfing, Bullish and Bearish Harami, Piercing Line and Dark Cloud Cover; and Morning Star, Evening Star, Three White Soldiers and Three Black Crows.

Does the candlestick skill make profitable trades?

The source gives no evidence of profitability. It applies simple pattern rules that ignore costs, risk management and market context, so you should backtest any signal yourself before relying on it.

Does the candlestick skill use the internet?

According to the skill's description, it does not use the network. It does run scripts, so read the included Python file before running it.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

This skill suits learners who want to see candlestick rules coded in pandas, and developers who need a starting pattern library for a prototype. It is not suited to making real trading decisions or to beginners who might trust the signal without backtesting, since the source offers no evidence that it is profitable.

Download the skill