EviACT: An Evidence-to-Action Framework for Agentic Program Repair

arXiv (Cornell University)

Abstract

LLM-based agents have moved automated program repair (APR) from fixed-context patch generation to interactive repository-level repair. However, existing agentic APR systems still struggle to use execution evidence to guide localization, patch generation, and validation. We propose EviACT (Evidence-to-Action), an agentic APR framework that coordinates three evidence-driven guardrails across repair stages. The retrieval scaffold grounds repair context, the compile gate filters invalid edits, and the test-driven gate checks target-test recovery before full regression. Across four benchmarks, EviACT improves resolve rate over the strongest reported comparable baselines by 1.6-6.0 percentage points and shows 70.1-88.6% lower reported per-bug API cost where baseline costs are available. Ablations and diagnostics suggest that these gains are associated with the coordinated evidence-to-action chain, making agentic APR more effective and efficient.

Authors 4

1. Qianru Meng
2. Xiao Zhang

   University of Groningen · Dialyse Centrum Groningen

   Affiliation as printed

   CLCG , University of Groningen Groningen , Netherlands

3. Zhaochun Ren Aachen

   Leiden University · University of Applied Sciences Leiden

   Affiliation as printed

   LIACS , Leiden University Leiden , Netherlands

4. Joost Visser Aachen

   Leiden University · University of Applied Sciences Leiden

   Affiliation as printed

   LIACS , Leiden University Leiden , Netherlands