Common Business Oriented Language (COBOL) is a programming language used to develop business applications that are widely adopted by financial, business, and government agencies. Due to its age, complexity, and declining number of COBOL developers, maintaining COBOL codebases is becoming increasingly challenging. In particular, the lack of documentation makes it difficult for new developers to effectively understand and maintain COBOL systems. Existing research utilizes large language models (LLMs) to explain the functionality of code snippets. However, COBOL presents unique challenges due to its architectural and syntactical differences, which often cause its code to exceed the token window size of LLMs. In this talk, we will present our multi-agent approach that works collaboratively to generate explanations for functions, files, and the overall project for COBOL projects. To support the code development, we will discuss our approach on applying LLM-based multiple agent structure for generating an initial code solution, iteratively refining the generated code and dynamically updating fixing strategies to improve the overall performance in code generation.