From Idea to Classroom in Days: Using "Vibe Coding" to Create a Programming Process Visualizer from IDE Activity Logs

Synthesis: Describes rapid development of a Thonny log visualizer using AI-assisted 'vibe coding' to make student programming processes visible to teachers. Piloted in a 160-student introductory programming course. Provides interactive timelines, session summaries, code-size graphs, and programming-process replays supporting teacher decision-making and Academic Integrity clarification.

Relevance to AI in Education: This paper contributes to the understanding of Automated Assessment, Personalized Learning, and Student Experience. The findings have implications for Adaptive Learning systems, Formative Assessment design, and the broader Edtech Platform landscape. Future work should explore how these results generalize across STEM Education and Higher Education contexts.

This research connects to the growing body of work on AI Literacy and Teaching, highlighting both the promise and limitations of AI tools in educational settings.

What this means for practice

• Instructors. Combine the student-level dashboard with cohort views to triage which logs deserve close reading, instead of inspecting submissions one by one.
• Instructors. Use process evidence — pauses, run-error cycles, time before the first run — to open a conversation about how a solution was built and to clarify academic-integrity concerns, not as an automated judgment.
• Instructors. Make log submission a course requirement from the start: the visualizer only works when students complete the mandatory tasks in Thonny and hand in the corresponding log files.
• Students. Replay your own session and check the code-size timeline before asking for help, so you can say where the time actually went.
• Designers. Plan for rapid iteration with teachers: this pilot produced concrete fixes — faster student switching, smoother replay, 15× and 30× speeds, red-flagged solving times under 10 minutes — within days of feedback.

Limitations

• The pilot ran in one "Introduction to Programming" course with 160 students and five teachers, and the authors state that this single-course context limits generalizability.
• Teacher feedback came from informal reviews and written comments collected during the course, with no control condition and no measured effect on student outcomes.
• The tool depends on a course workflow that requires Thonny and log-file submission, so it does not transfer to courses without that logging infrastructure.
• The views are built to prompt follow-up discussion and manual review rather than to serve as automated judgments, and time-window filtering of logs was requested but not yet implemented.

Citation

Heidi Taveter, Marina Lepp (2026). From Idea to Classroom in Days: Using Vibe Coding to Create a Programming Process Visualizer from IDE Activity Logs. FIE 2026 (IEEE Frontiers in Education Conference).