Building an automated stock analysis system without writing a single line of code feat. Claude CLI

[Hodu's AI Lab] Building a Practical Stock Analysis System with Claude CLI

You must use the Sonnet 4.6 model to ensure you do not exceed the usage limits.

The free Google API has been discontinued as of March 2026.

⚠️ Investment Disclaimer (Disclaimer)

- This course covers the methodology for anyone to create a system that quickly analyzes stocks through Vibe Coding, and was created for the purpose of teaching how to build an automated analysis system using widely used strategies in the current market, such as closing price betting, VCP pattern analysis, Smart Money screening, and ML prediction. It does not recommend the buying/selling of specific stocks or guarantee investment returns.

- The returns, win rates, and backtest results shown in the lecture are simulations based on sample and test data, and past performance does not guarantee future returns.

- The stock names, scores, and ratings mentioned during the lecture are examples for system demonstration purposes and are not investment recommendations.

- This lecture does not constitute investment advisory or investment management services under the Financial Investment Services and Capital Markets Act.

- All responsibility for profits and losses resulting from actual investments lies with the investor themselves, and the course creator bears no legal responsibility for the student's investment results.

-AI and data analysis tools are merely means to assist in investment judgment; final investment decisions must be made based on your own judgment and responsibility.

[Curriculum Guide]

Currently, Parts 1 and 2 are available, and one part will be updated sequentially every week. (A total of 10 parts are scheduled for completion.)

Part 1: Installing Claude Code and Setting Up the Development Environment (Using Antigravity)

Part 2: Python Basics Practice (Building a calculator and crawler by instructing AI)

Part 3: Lecture update completed (2/23)
Part 4: Lecture update completed (2/28)

Special Lecture: Telegram Integration Completed (3/5)

Part 5: Lecture Update Completed (3/8)

Part 6: Lecture update completed (3/18)

Part 7: Next.js_Dashboard Completed (3/27)

Part 8: Automation_Scheduler Completed (4/7)

"You don't need to know coding. Just learn how to 'properly' instruct the AI."

Hello, I am Hodu, the creator of the YouTube channel <Hodu's AI Analysis Lab>. While running the channel, there is one question I have received most frequently and consistently from my subscribers.

"Can I also build the stock analysis system you use?" "I don't know anything about coding; is there a lecture that teaches me from start to finish?"

I created this course to quench that thirst of yours. These are not just simple examples. I have captured, without filter, the entire process of building the 'Automated Stock Analysis System' that I use for actual trading every morning from scratch using Claude CLI.

Why 'Claude CLI'?

Don't be fooled by those '5-minute coding' videos on YouTube anymore. You cannot build complex systems through the manual labor of copying and pasting code line by line from a web chat window.

In this course, we use the latest AI development tool, Claude CLI. You will experience true automated development, where the AI directly creates project files, modifies code, and even fixes errors within the terminal environment.

Key Course Point: Real-world Full-stack Workflow

In lessons 1 to 3, you will experience creating a stock program exactly as shown using the prompts provided in my files.

From Lesson 4 onwards, you will begin learning how to use prompts and the "Why" behind writing them this way.

✅ Real Data Handling Building a pipeline to collect and process real-time price data for all 2,500 stocks

✅ Advanced Logic Implement a scoring engine that crawls news data, performs sentiment analysis using AI, and quantifies your own investment strategies

✅ Full Automation From building an API server to web dashboard visualization, and even Telegram reports automatically delivered every morning

After taking this course

Instead of simply typing along with the code, you will gain the 'ability to design like a developer.'

• Scalability: When new analysis indicators or strategies come to mind, you will be able to add them with confidence.

• Structuring: You will gain the power to conquer complex systems by breaking them down step by step.

• Adaptability: You can build any data analysis system, whether for real estate or cryptocurrency, not just stocks, using the same structure.

The fastest way to turn your ideas into reality, start now with Hodu using the Claude CLI.

👉 Go to Hodu's AI Analysis Lab YouTube Channel: https://www.youtube.com/@두두감자

After taking the course, you will be able to create results like this

Learning Content

Notes before taking the course

🛠️ Hands-on Practice Environment Guide

Check before you start! This course was filmed in a macOS environment, but Windows and Linux users can also take it without any issues. A standard office or development laptop without a GPU (graphics card) is sufficient.

1️⃣ Operating System (OS)

The course is conducted based on macOS, but 99% of the content is the same regardless of the OS.

Differences between operating systems (such as scheduler settings) will be kindly explained separately within the lecture.

OS Recommended Version Remarks macOS macOS 13 (Ventura) or higher Standard environment for the course (utilizes native features like launchd) Windows Win 10/11 + WSL2

After installing WSL2 (Ubuntu), you can follow the same hands-on practice.

(For the scheduler, use Windows Task Scheduler)

LinuxUbuntu 22.04 or higher. Fully compatible using cron, etc.

2️⃣ Essential Tools and Software

Since AI handles the coding, you only need to prepare the tools to give instructions (Prompts).

Core Tool: Claude Code (CLI) Paid ($20/month~)

The core engine of Vibe Coding. Backend. Python (3.9 or higher) Free. Data analysis, crawling, and server operation.

Frontend Node.js (18 or higher)FreeRuns Next.js-based dashboard screens

Editor VS Code FreeCode verification and terminal execution (Optional)

3️⃣ External APIs (No additional cost)

Instead of using complex paid data services, we maximized cost-effectiveness by utilizing free crawling and API free tiers.

• Google Gemini API: News AI summarization and sentiment analysis (utilizing free tier)

• Naver OpenAPI: News search (25,000 requests per day for free)

• Telegram Bot API: Sending buy/sell notifications (completely free)

• 🚫 Stock price data: Handled via web crawling without a separate paid API.

4️⃣ Recommended PC Specifications

Since you won't be running machine learning training yourself, a high-spec PC is not required.

• CPU: Dual-core or higher (Recommended: Apple M1~M3 or Intel i5 or higher)

• RAM: 8GB (Recommended: 16GB or higher)

• Disk: 10GB or more of free space

• Internet: Stable wired/Wi-Fi environment required

📚 Learning Materials Guide

To ensure the successful system construction for all students, we provide both the exclusive prompts and source code.

📦 Provided Materials

1. Prompt Collection (Notion file provided) ✨ (Core Material)

   • This is the alpha and omega of the course. We provide over 120 commands to be entered into Claude Code for each part, along with a troubleshooting guide. Experience building a system simply by copying and pasting.

2. Lecture Notes & Slides

   • It includes the full architecture diagram, a summary of core concepts, and a checklist.

3. Completed Source Code

   • This is a reference (answer key) you can consult in case you get lost.

💡 Tips for Using the Materials

• Try creating it yourself: The provided complete code is for reference only. Even if the code you create by entering prompts yourself has slightly different variable names, if the functionality works, that is the correct answer.

• Maintenance Guide: If the code stops working because the target crawling site (such as Naver Finance) has been updated, simply provide the error message to Claude Code, and it will fix it for you. (This method is also covered in the lecture!)

🎯 Prerequisites & Learning Guide

"Is it okay if I don't know how to code?" 👉 Yes, it is possible!

This course is not about becoming a developer, but about using AI as a tool to achieve the desired result (a stock analyzer).

• Programming experience (Not required): Claude writes the code using the 'Vibe Coding' method.

• Python/Terminal knowledge (Not required): I will guide you step-by-step starting from basic terminal usage (Part 0).

• Stock Basics (Recommended): Knowing basic terms like moving averages and trading volume will help you understand investment strategies.

🚀 Recommended Learning Roadmap

1. Type it yourself: Don't just watch with your eyes. You must experience the process of entering prompts yourself and seeing the AI spit out code for it to truly become yours.

2. Follow the order: The structure is designed to build up step-by-step from Part 1. Skipping parts may lead to errors.

3. Don't be afraid of errors: If an error occurs, don't panic; copy the error message and ask Claude. That is the core skill you are learning in this course.

4. Reviewing: In particular, Part 4 (Closing Price Betting Engine), Part 7 (Dashboard), and Part 9 (Backtesting) contain deep content. I recommend taking your time and reviewing them 1–2 times.

📢 Q&A & Updates

Q&A and Feedback

• Please use the [Inflearn Q&A board] for questions.

• When asking a question, please include the entire error message and the OS you are using to receive a much faster and more accurate response.

Lecture Update Policy

• Responding to Site Structure Changes: Prompts and code are updated when the structure of a crawled site changes.

• Claude Code Updates: We check and announce compatibility during major tool updates.

• Copyright Notice: Please use the prompts and materials from this lecture for the student's personal learning purposes only. (Unauthorized distribution/sharing is prohibited)

⚠️ Investment Precautions (Disclaimer)

The closing price betting strategy, VCP pattern analysis, and automated trading system covered in this lecture were created for the purpose of educating on system construction methodology.

Following the course content does not guarantee profits, and the student bears full responsibility for all gains and losses resulting from actual investments. Please use this wisely as a supplementary tool utilizing AI and data.