Learning When and How to Intervene: A Hindsight-Distilled Sentinel for Coding Agents

Excerpt

arXiv:2609.39957v1 Announce Type: cross Abstract: Coding agents solve repository-level tasks through sequences of actions, where a single erroneous action can misdirect subsequent decisions and increase recovery costs. Existing approaches use execution feedback for recovery or specialized checks to block errors, but deciding before execution whether intervention will benefit eventual task completion remains challenging. To address this challenge, we propose HiSentinel, a hindsight-distillation f