Is your feature request related to a problem? Please describe.
According to @jiwenc-nv's rough estimation, a big portion of code (>90%) is either generated or coded with with AI assistant. However, Isaac Teleop currently does not provide any guidelines on coding style, design pattern, best practices. We are spending a great a mount of time on human-based code review, which does not scale. Worse: when things need to move fast, quality slips and AI slops accumulates.
Describe the solution you'd like
- Setup proper guidelines for AI to follow.
- Setup automated agentic workflow to check PRs and give actionable feedbacks that aligns with our guidelines.
- Continuously monitoring the code consistency and make refactors.
Describe alternatives you've considered
- Manual review -> doesn't not scale.
- No review -> quality slips and code base become more and more chaotic.
Additional context
TBA
Is your feature request related to a problem? Please describe.
According to @jiwenc-nv's rough estimation, a big portion of code (>90%) is either generated or coded with with AI assistant. However, Isaac Teleop currently does not provide any guidelines on coding style, design pattern, best practices. We are spending a great a mount of time on human-based code review, which does not scale. Worse: when things need to move fast, quality slips and AI slops accumulates.
Describe the solution you'd like
Describe alternatives you've considered
Additional context
TBA