auto config env

IMPORTANT

This plugin uses pip to install packages within a Conda environment, defaulting to the base environment. After the plugin is deployed, please select the corresponding Conda environment or create a new environment from the bottom-right corner.

Features

1. Complete Imports: Completes imports in a single Python file.
2. Auto Config: Utilizes a large language model to automatically install Python packages with pip.
3. Environment Debug: Automatically runs the Python file. If errors occur, the large language model suggests how to set up the environment and executes the returned commands. The debug loop runs up to 3 times.
4. Generate requirements.txt (Explore Folder). Warning: This operation will generate and replace the requirements.txt file in the selected folder for .py files in all subdirectories.

Deploy

git clone <repository-url>
cd <project-directory>
npm install 
code .

run it with F5 in vscode.

Requirements

• OpenAI key (Current model: gpt-3.5-turbo-1106, you can choose prefered model in Settings, but it might lead to errors due to different api.)
• Conda

Configuration

Open VSCode.

Navigate to File > Preferences > Settings (or press Ctrl + , / Cmd + ,). Search for "auto_config_env" and enter your API key in the input box.

Notes

• The extension is designed for Ubuntu and not tested on Windows.

Extension Settings

This extension contributes the following settings:

• "auto_config_env.openaiApiKey": "Your OpenAI Key"
• "auto_config_env.gpt_model": "gpt-3.5-turbo-1106" or other models(without tests)

Known Issues

Let me know if there's anything else you'd like to add or change!

TODO

Support other LLMs like claude.