Skip to content
| Marketplace
Sign in
Visual Studio Code>Other>Coding AssistantNew to Visual Studio Code? Get it now.
Coding Assistant

Coding Assistant

csci435CodingAssistant

|
1,620 installs
| (0) | Free
The Coding Assistant is a VSCode extension that utilizes Natural Language Processing (NLP) in the form of Large Language Models (LLMs) and Transformers to improve the software development experience within the IDE.
Installation
Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter.
Copied to clipboard
More Info

Coding Assistant

Description

The Coding Assistant is a VSCode extension that utilizes Natural Language Processing (NLP) in the form of Large Language Models (LLMs) and Transformers to improve the software development experience within the IDE. This extension’s features include code autocompletion and git commit message generation. The Coding Assistant aims to expedite the development process while providing general quality of life improvements to developers.

Features:

  • Code autocompletion
  • Auto git commit message generation

Getting Started

  1. Create a Hugging Face account at https://huggingface.co/. Once logged in, create a read/write access token by accessing your account settings and navigating to “Access Tokens” on the sidebar Preview image Preview image Preview image
  2. Paste the token into the extension setting correlated to API key Preview image Preview image
  3. Search “inline” in Settings and check the “Inline Suggest” box Preview image

Usage

  • Code completion suggestions will be generated via typing code and can be cycled through with Alt + [ or ]. Accept a suggestion with the Tab key.

Preview image

(Gif taken from Code Clippy)

  • To generate git commit messages, run git diff in the command line to fetch the changes made since the last commit, and paste them into a text file. Highlight the entire block, then right click and select the Git Commit Message option. The generated commit message should appear above the git diff block.

Preview image

Future Work and Improvements

  • In the future, more can be added to this extension to make it more useable and more helpful:
    • Add in other two features from project* (Text to Code, Code Commenter)
    • Implement other possible features (code clone detection, auto generation of unit tests)
    • Find and/or create more representative datasets on which to train models
  • Also, many aspects of the extension can be improved, such as:
    • Enabling hosting for feature APIs
    • Speed and accuracy of responses
    • Refactoring to further condense the code base
    • More effective models to use for features

*To learn more about the project, visit our website: https://github.com/WM-SEMERU/csci-435_coding_assistant

Authors

Sarah Wille, Isabella Wu, Jimmy Yao, Effie Zhang, Robert Zhang

Acknowledgements

  • Code Clippy
  • Models Used:
    • Autocomplete: GPT-CC - Open source Github Copilot model based on GPT-3, trained on publicly available code from GitHub repositories meeting specific criteria.
    • Auto git commit message generation: CodeTrans - Pre-trained using t5 large model architecture and fine-tuned on the tokenized git commit message generation task.
  • Contact us
  • Jobs
  • Privacy
  • Manage cookies
  • Terms of use
  • Trademarks
© 2025 Microsoft