Cloud-Based System for Source Code Analysis of Microservices with LLM Agents
Oleh Chaplia, Halyna Klym · 2024
This paper presents a cloud-based system for microservice code analysis that utilizes LLM agents as a foundation. The system is designed to receive the code updates, analyze the code, and provide results to the user. Users can obtain information about the microservice, code summaries, code reviews, reliability checks, and improvement recommendations. The history of changes allows for tracking the evolution of microservices and identifying the issues when they appear. The prototype of the proposed system was implemented and tested. The results show the systems’ usability and provide valuable insights. This research underscores the potential of combining LLM agents with cloud technologies, offering a scalable microservice solution that paves the way for future innovations in automated code analysis and software engineering.