SOBO: A Feedback Bot to Nudge Code Quality in Programming Courses
summary
SOBO is an automated GitHub bot that analyzes Java commits with SonarSource rules (curated from 662 down to actionable ones), opens a persistent GitHub issue with a clickable file location, an example fix, and simple commands. It is deployed to 130+ students across ~1,600 repositories, runs static analysis (SonarSource/SpotBugs style), and lets students stop/restart feedback. Quantitative results are mixed—more violations are introduced than fixed overall, but several students stop repeating magic number violations after early nudges; retention and template tuning remain work.
AI-written summary · about
chapters (18)
- 0:00SOBO introduction
- 1:31Programming course context
- 3:38Code feedback challenges
- 5:01Bots in education
- 7:51Code quality awareness need
- 9:15SOBO the feedback bot
- 12:30SOBO’s quality rules
- 14:38Feedback channel choice
- 15:19Automated feedback explained
- 17:29SOBO interactive features
- 18:58SOBO live demo
- 25:53Experiment results
- 28:42Key takeaways
- 30:08Lessons for bots
- 31:53Long-term learning impact
- 34:01Future improvements
- 36:07Current status
- 36:51Q&A
related talks
30mSCORE Bot: Shift Left, at Scale!
16mShift Left with GitHub Copilot: Enforcing Code Quality Before the Commit
40mLet a bot deal with your static analysis warnings backlog
46mIn search of Clean Code with Sonar, 20,000 leagues under an ocean of code!
08mHow to Recommend Tools for Finding and Fixing Software Errors
13mHow to Improve AI-Generated Code with SonarSweep
30mAn Extensible Feature-Oriented Approach for Fine-Grained Code Quality Analysis
43mCode Review Tools