What problem does it solve?
It prevents mistakes when end users call NVIDIA cuOpt by guiding you to clarify requirements, verify your environment, follow correct setup/usage expectations, and confirm solver results before acting on them.
Core Features & Use Cases
- Clarify before building: Prompts you to confirm language/interface, problem type, constraints, and output format to avoid implementing the wrong model.
- Validate inputs and assumptions: Helps you document what data you used (user data vs. synthesized), state assumptions explicitly, and confirm variable/constraint expectations.
- Verify environment and execution: Instructs how to check the correct cuOpt installation path by interface (Python, C library, REST server, CLI) and how to ask permission before running verification commands.
- Guardrails for correct output: Enforces result checking (solver status and objective value visibility) and calls for a post-correction verification step when changes are needed.
Quick Start
Use the cuopt-user-rules skill to ask cuOpt for help only after confirming your problem formulation, constraints, and interface details, then verify the solver status and objective value prominently in the result summary.
