Accessibility Heuristics for Vibe Coding Interfaces

Shalini Madan, Sreelakshmi Surabiyil Bindhu, Venkatesh Potluri.

To appear in Posters at the ASSETS 2025 Symposium on Computers and Accessibility (ASSETS 2025).

Abstract

AI coding tools are transforming programming, shifting it from a highly editorial process into a conversational activity. Popular Vibe coding tools such as Replit and Cursor integrate natural language interfaces with traditional development environments. While these tools promise simplicity, increased productivity, and automation, they also introduce new accessibility challenges for blind or visually impaired (BVI) developers. A systematic identification of these accessibility challenges necessitates comprehensive guidelines that account for the complex interactions in these tools. To address this need, we develop accessibility heuristics to assess the accessibility of AI conversational programming tools. Our heuristics combine web accessibility guidelines, best practices to design conversational interfaces, and accessibility needs specific to BVI developers. Our evaluation of three widely used conversational programming tools shows that most accessibility challenges arise from insufficient feedback and access to the various actions and output of the tools, complex keyboard interactions, and poor focus management.