10th U6+ Consortium of African Universities International Conference

Traditional Computer Science (CS) capstone project supervision often suffers from a "black box" syndrome, where the development process remains opaque to supervisors until the final submission. This disconnect frequently leads to severe student procrastination, poor architectural planning, and a lack of alignment with industry-standard engineering practices such as Continuous Integration (CI) and collaborative peer review. This study introduces a cloud-native, infrastructure-agnostic supervision model designed to bridge the industry-academia skills gap. The framework changes the focus of teaching from the final software product to a clear, milestone-driven engineering process. By employing GitHub Classroom as a virtual laboratory environment, the Framework implements four mandatory technical gates: architectural validation using Diagrams-as-Code (Mermaid.js), automated quality assurance via CI/CD (GitHub Actions), social coding audit through pull requests, and deployment and performance verification. To measure progress, we propose and validate the Weighted Technical Proficiency Rubric (WTPR), which maps these GitHub-native artifacts to the IEEE SWEBOK v4.0 professional standards. As an infrastructure-focused design paper, we demonstrate that the framework provides a scalable, zero-cost blueprint for high-fidelity supervision. Initial logic validation suggests that the framework successfully mitigates "big-bang" submissions by enforcing continuous transparency and provides students with a verifiable digital engineering portfolio to enhance employability.