CodeRunner Agent: Integrating AI Feedback and Self-Regulated Learning to Support Programming Education

Abstract

The emergence of Large Language Model (LLM) tools has revolutionized programming education, demonstrating considerable efficacy in providing instant, personalized feedback. However, most existing LLM-based tools focused on direct coding assistance, and neglected the cultivation of self-regulation skills. Moreover, many of these tools operate independently from institutional Learning Management Systems, which creates a significant pedagogical disconnect that limits the ability to leverage contextual learning materials and exercises for generating tailored, context-aware feedback. To address these challenges, we developed CodeRunner Agent, an LLM-based programming assistant. CodeRunner Agent enhances students' self-regulated learning by providing strategy-based AI feedback and embedding within Moodle system. Additionally, it empowers educators to customize AI-generated feedback by incorporating detailed context from lecture materials, programming questions, student answers, and execution results. This integrated approach, emphasizing self-regulation skill development with contextual awareness, offers promising avenues for data-driven enhancements in programming education with AI.