Answers are grounded in this paper and cite their sources.
Capability audit demonstrates functional correctness and boundary compliance in introductory C++ tasks, suggesting critical limitations in automated programming assistance.
Companion material associated with the work "A Course-Anchored Capability Audit of GitHub Copilot for Introductory Programming: Specification and Code Generation". This deposit provides the artifacts required to inspect and reproduce the reported course-anchored capability audit of GitHub Copilot. It includes the 14 introductory C++ programming tasks used in the study, instructor-authored reference implementations and specifications, the original Spanish code-generation prompts together with English translations, the operational scoring rubrics, adjudicated slot-level scoring records for specification and code generation, and supplementary figures and numerical summaries. The study evaluates generated artifacts rather than student learning outcomes. The materials therefore support auditing of GitHub Copilot's availability, non-repetition, specification accuracy, functional correctness, curricular-boundary compliance, redundancy, correction effort, and variability across repeated runs. One run corresponds to one complete execution of a fixed prompt requesting up to ten candidate alternatives. The numerical entries in the preserved scoring records are unchanged from the original study records.
No takes yet. Share an insight, caveat, or question.
Flores-Martín et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: