Developed and delivered the System Prompt Feature for the chat script in the ml-explore/mlx-lm repository, enabling users to specify guiding prompts that enhance the evaluation of fine-tuned chat models by providing configurable context. The implementation focused on robust Python back end development, integrating automated unit tests to ensure reliable functionality and seamless integration with existing workflows. All changes were tracked through clear Git-based commit history, supporting traceability and review. This work improved the reliability of experiment results and streamlined evaluation processes, demonstrating proficiency in Python, test automation, and CI-friendly development practices within a collaborative open-source environment.