AI Prompts Trigger a Fundamental Reassessment of Software Libraries, Reshaping Dependency Management

Artificial intelligence is rapidly transforming software development, particularly the role and perceived value of external libraries. A growing trend, dubbed ‘vibe coding,’ sees developers opting to rip out and rewrite code that was once provided by smaller, easily understood libraries (e.g., is-odd, left-pad) using AI prompts. This phenomenon is driven by concerns over dependency bloat, supply chain security risks, and the desire for greater control and understanding of the codebase. Paradoxically, despite AI’s ability to generate these simple solutions, downloads for such basic libraries appear to be increasing, suggesting a surge in less experienced developers entering the field and still relying on established package installation patterns.

This evolving landscape also impacts more complex projects. While AI still struggles with foundational or highly nuanced systems like FFmpeg or the Bun runtime, it excels at providing context and generating solutions for well-understood problems. This leads to the emerging concept of ‘prompts as the new libraries,’ where developers share AI prompts to implement common functionalities rather than distributing traditional packages, fostering direct integration and local control over the code. Companies like Anthropic are demonstrating this shift strategically: deciding to internally rewrite Ink (a React-for-CLI library) due to its ease of re-implementation and maintenance benefits, while acquiring Bun due to its strategic importance as a core dependency that wraps their entire project. The value of having code directly within a repository is escalating, as AI agents perform significantly better with in-context code, as exemplified by projects like Shadcn UI, which promotes copy-pasting components into the codebase. This fundamental re-evaluation of dependency adoption calculus – lower perceived implementation difficulty via AI versus higher perceived risk of external dependencies – is set to redefine how projects are built and maintained.