Agentic Generation of AST Transformation Rules for Fixing Breaking Updates

2026-06-23

AI Summary

This work addresses the challenge posed by breaking API updates in third-party libraries, which frequently cause downstream projects to fail compilation, and existing patch-generation approaches often produce fixes that are not reusable across projects. To overcome this limitation, the authors propose BigBag, a novel framework that, for the first time, automatically generates transferable, transformation-based repair rules tailored to breaking API changes. BigBag integrates large language models with an AST transformation engine (leveraging Spoon and JavaParser) and employs a multi-agent collaboration strategy to synthesize structured, executable, API-level repair logic. Evaluated on the BUMP benchmark, BigBag achieves a 94.3% compilable transformation rate and a 78.6% overall repair rate, with 33.3% of fixes successfully transferred across projects—exceeding 80% in unified invocation scenarios—thereby substantially surpassing the constraints of project-specific patches.

Abstract

Modern software projects depend on third-party libraries that evolve continuously, introducing breaking API changes that prevent client code from compiling after a dependency update. When the same library update breaks multiple projects, existing repair approaches generate project-specific patches that cannot be reused, requiring each affected project to be repaired independently. We present BigBag, an agentic framework that generates fixing transformations: structured, executable programs that encode the repair logic at the API level and transfer to any client broken by the same update. We evaluate BigBag on 157 compilation failure breaking dependency updates from the BUMP benchmark, across eight configurations combining four large language models and two AST transformation engines (Spoon and JavaParser). The best configuration achieves a compilable transformation rate of 94.3% and a fix rate of 78.6%. Generated transformations transfer across projects, achieving a cross-project fix rate of 33.3% overall and 80% or above for breaking updates where all clients invoke the affected API element uniformly. These results show that agentic generation of reusable fixing transformations is a viable approach to scalable repair of breaking updates.

Problem

• breaking updates
• API changes
• code repair
• AST transformation
• reusable patches

Innovation

• agentic repair
• AST transformation
• breaking API updates
• cross-project transferability
• reusable fixes