How ai agents change software libraries with dependency transformations
Agentic Dependency Transformations: Lessons from an Exploratory Case Study of Flask
Software Engineering
Summary
Modern software often uses code from others, called dependencies, which can cause problems when those external pieces change or break. The authors studied how AI coding agents can change how software uses these dependencies, sometimes by replacing them with their own copy of the code. They looked at the popular Flask software to see how these AI agents find and change dependencies step-by-step. From this, they learned five important lessons about how AI agents handle dependencies and suggested ideas for safely managing these changes in the future.
What this means in practice
- •For software maintainers: Use AI agents to discover and gradually replace or modify software dependencies to reduce maintenance risks.
- •For ai tool developers: Design AI coding tools that respect transformation contracts and validation criteria when automating dependency changes.
Tested on one dataset.
Authors
Kaya Kitada, Yoshiki Higo, Raula Gaikovina Kula
Abstract
Modern software relies heavily on third-party libraries, requiring maintainers to balance the benefits of reuse against dependency-related maintenance and risk. We define dependency transformation as an emerging approach in which an AI coding agent alters how a repository handles external dependencies. Local reimplementation is one transformation strategy that recreates required functionality as local code; dependency transformation is broader and may include other strategies. In this emerging-results paper, we report an exploratory case study of Flask to explore dependency transformation strategies. We design experiments that show how available context shapes an agent's dependency discovery and iterative transformation of dependencies. The case yields five preliminary lessons. These lessons and the five-part research agenda position dependency transformation as an emerging research direction and motivate explicit transformation contracts, provenance constraints, preservation requirements, and independent validation criteria.