zhongerxin/self-organizing-life

The AI Python Code Interpreter lets you describe a task in plain English and get working Python code that actually runs, without writing the code yourself. You type something like "read a CSV file, calculate averages, and make a bar chart," and the system writes the code, installs any missing third-party packages, runs it, and hands you the results. Under the hood, it sends your request to Claude, which generates the Python code along with a plain-language explanation of what it does. The system spots any external packages the code needs and installs them automatically. It then runs the code inside an isolated virtual environment, which keeps your main system safe and clean. After execution, you get the output, any error messages, and how long it took. The main audience is people who want Python's power without writing or debugging it from scratch. A startup founder could ask it to analyze a spreadsheet and chart the results. A product manager could ask it to count lines of code across project files. A researcher could fetch weather data from an API and summarize it. You can use it through a command-line prompt, an interactive session, or a web API that other software can call. A few design choices stand out. Code runs in a throwaway virtual environment with execution time limits, so an infinite loop or bad script won't tie up your machine. Temporary files are cleaned up automatically. The project also exposes two API endpoints: one that generates and optionally runs code, and another that just runs code you provide.