AICodeMentor leverages Large Language Models to function as an intelligent coding companion. Developers feed code snippets or entire projects, and the agent performs in-depth analysis to detect potential bugs, coding style violations, and security vulnerabilities. It then generates constructive feedback, suggests optimizations, and produces unit test scaffolding. Additionally, it supports interactive question-and-answer sessions for clarifying complex logic or library usage. By integrating easily via pip installation or direct GitHub clone, AICodeMentor can be invoked in CI pipelines, local development, or educational settings. Its modular architecture allows customization of prompts and LLM providers, delivering a tailored mentoring experience that scales from beginners learning syntax to seasoned engineers seeking code refinement.