This research introduces a novel framework that automates code debugging using large language models, improving development efficiency and code quality.

• Combines LangGraph (for orchestration), GLM4 Flash (LLM), and ChromaDB in a four-step iterative workflow
• Enables automated detection and fixing of software bugs without human intervention
• Improves engineering productivity by reducing manual debugging time
• Creates more reliable and scalable software development processes

For engineering teams, this represents a significant advancement in automating tedious debugging tasks, allowing developers to focus on higher-value creative work while improving overall code quality.

Empirical Research on Utilizing LLM-based Agents for Automated Bug Fixing via LangGraph