Agentic Automated Program Repair
What it is
An autonomous LLM agent built on LangGraph that repairs buggy code by executing it, parsing compiler errors, and iteratively rewriting the source until it passes validation. It targets specific defect types like off-by-one and semantic errors, claiming an 86% fix rate on the QuixBugs benchmark.
The problem it solves
Manual debugging of simple, single-line logic errors (like off-by-one or semantic slips) is time-consuming and repetitive for developers.
What you could use it for
- Automatically fixing logic errors in Python scripts
- Repairing off-by-one errors in array or loop logic
- Validating and correcting single-line semantic bugs in codebases
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.