Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

CodingGenie: A Proactive LLM-Powered Programming Assistant

Sebastian Zhao, Alan Zhu, Hussein Mozannar, David Sontag, Ameet Talwalkar, Valerie Chen

TL;DR

CodingGenie introduces a proactive LLM-powered coding assistant integrated into VSCode via an open-source Continue extension. It generates context-aware, diverse suggestions across six categories based on the current code context and a user-provided task description, with configurable suggestion types. The paper evaluates utility across personal projects, industry tickets, and debugging school tasks, showing that customization boosts relevance, and provides a cost analysis indicating manageable overhead that can be tuned. Overall, CodingGenie demonstrates how proactive assistants can be integrated into real-world coding workflows and serves as a platform for further research into proactive development tools.

Abstract

While developers increasingly adopt tools powered by large language models (LLMs) in day-to-day workflows, these tools still require explicit user invocation. To seamlessly integrate LLM capabilities to a developer's workflow, we introduce CodingGenie, a proactive assistant integrated into the code editor. CodingGenie autonomously provides suggestions, ranging from bug fixing to unit testing, based on the current code context and allows users to customize suggestions by providing a task description and selecting what suggestions are shown. We demonstrate multiple use cases to show how proactive suggestions from CodingGenie can improve developer experience, and also analyze the cost of adding proactivity. We believe this open-source tool will enable further research into proactive assistants. CodingGenie is open-sourced at https://github.com/sebzhao/CodingGenie/ and video demos are available at https://sebzhao.github.io/CodingGenie/.

CodingGenie: A Proactive LLM-Powered Programming Assistant

TL;DR

CodingGenie introduces a proactive LLM-powered coding assistant integrated into VSCode via an open-source Continue extension. It generates context-aware, diverse suggestions across six categories based on the current code context and a user-provided task description, with configurable suggestion types. The paper evaluates utility across personal projects, industry tickets, and debugging school tasks, showing that customization boosts relevance, and provides a cost analysis indicating manageable overhead that can be tuned. Overall, CodingGenie demonstrates how proactive assistants can be integrated into real-world coding workflows and serves as a platform for further research into proactive development tools.

Abstract

While developers increasingly adopt tools powered by large language models (LLMs) in day-to-day workflows, these tools still require explicit user invocation. To seamlessly integrate LLM capabilities to a developer's workflow, we introduce CodingGenie, a proactive assistant integrated into the code editor. CodingGenie autonomously provides suggestions, ranging from bug fixing to unit testing, based on the current code context and allows users to customize suggestions by providing a task description and selecting what suggestions are shown. We demonstrate multiple use cases to show how proactive suggestions from CodingGenie can improve developer experience, and also analyze the cost of adding proactivity. We believe this open-source tool will enable further research into proactive assistants. CodingGenie is open-sourced at https://github.com/sebzhao/CodingGenie/ and video demos are available at https://sebzhao.github.io/CodingGenie/.

Paper Structure

This paper contains 18 sections, 1 equation, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Tool Overview
  3. Generating Proactive Suggestions
  4. What kinds of suggestions?
  5. What context is included in the prompt?
  6. How can users customize suggestions to their needs?
  7. Interacting with Proactive Suggestions
  8. Interface.
  9. Refresh Logic.
  10. Implementation Details
  11. Evaluation
  12. Evaluation of Utility
  13. Personal Project.
  14. Industry Ticket Completion.
  15. Debugging School Assignment.
  16. ...and 3 more sections

Figures (2)

  • Figure 1: Overview of the CodingGenie tool. The left side of the diagram depicts the elements in a developer's workspace that make up the prompt used to generate proactive suggestions, and the right side depicts the user interface with proactive suggestions (shown in the light blue box) upon a successful generation.
  • Figure 2: System design diagram for communication between the GUI, extension, and editor components of Continue to implement proactive suggestions.