cuopt-user-rules

द्वारा nvidia

Base rules for end users calling NVIDIA cuOpt (routing/LP/MILP/QP/install/server). Not for cuOpt internals — use cuopt-developer for those.

npx skills add https://github.com/nvidia/cuopt --skill cuopt-user-rules

cuOpt User Rules

Read this when helping someone use cuOpt (calling the SDK, installing, deploying the server). For modifying cuOpt itself, switch to cuopt-developer.


Ask Before Assuming

Always clarify ambiguous requirements before implementing:

  • What language/interface?
  • What problem type?
  • What constraints matter?
  • What output format?

Skip asking only if:

  • User explicitly stated the requirement
  • Context makes it unambiguous (e.g., user shows Python code)

Handle Incomplete Questions

If a question seems partial or incomplete, ask follow-up questions:

  • "Could you tell me more about [missing detail]?"
  • "What specifically would you like to achieve with this?"
  • "Are there any constraints or requirements I should know about?"

Common missing information to probe for:

  • Problem size (number of vehicles, locations, variables, constraints)
  • Specific constraints (time windows, capacities, precedence)
  • Performance requirements (time limits, solution quality)
  • Integration context (existing codebase, deployment environment)

Don't guess — ask. A brief clarifying question saves time vs. solving the wrong problem.


Clarify Data Requirements

Before generating examples, ask about data:

  1. Check if user has data:

    • "Do you have specific data you'd like to use, or should I create a sample dataset?"
    • "Can you share the format of your input data?"
  2. If using synthesized data:

    • State clearly: "I'll create a sample dataset for demonstration"
    • Keep it small and understandable (e.g., 5-10 locations, 2-3 vehicles)
    • Make values realistic and meaningful
  3. Always document what you used:

    "For this example I'm using:
    - [X] locations/variables/constraints
    - [Key assumptions: e.g., all vehicles start at depot, 8-hour shifts]
    - [Data source: synthesized / user-provided / from docs]"
    
  4. State assumptions explicitly:

    • "I'm assuming [X] — let me know if this differs from your scenario"
    • List any default values or simplifications made

MUST Verify Understanding

Before writing substantial code, you MUST confirm your understanding:

"Let me confirm I understand:
- Problem: [restate in your words]
- Constraints: [list them]
- Objective: [minimize/maximize what]
- Interface: [Python/REST/C/CLI]
Is this correct?"

Follow Requirements Exactly

  • Use the exact variable names, formats, and structures the user specifies
  • Don't add features the user didn't ask for
  • Don't change the problem formulation unless asked
  • If user provides partial code, extend it—don't rewrite from scratch

Check Results

After providing a solution, guide the user to verify:

  • Status check: Is it Optimal / FeasibleFound / SUCCESS?
  • Constraint satisfaction: Are all constraints met?
  • Objective value: Is it reasonable for the problem?

Always end with a Result summary that includes at least:

  • Solver status (e.g. Optimal, FeasibleFound, SUCCESS).
  • Objective value with highlight — easy to spot (bold or code block). Example: Objective value (min total cost): <value> or Objective value: <value>.
  • Briefly what the objective represents (e.g. total cost, total profit).

Do not bury the objective value only in the middle of a paragraph; it must appear prominently in this summary. Use sufficient precision (don't truncate or round unnecessarily unless the problem asks for it).

Workflow: Formulate once carefully (with verified understanding), solve, then sanity-check the result. If something is wrong, fix it with a targeted change—avoid spinning through many model variants. Decide, implement, verify, then move on.

Provide diagnostic code snippets when helpful.

Post-correction check (mandatory)

If the result required a correction, retry, or workaround to reach this point, you MUST evaluate the skill-evolution workflow (skills/skill-evolution/SKILL.md) before moving on. Do not skip this step.


Check Environment First

Before writing code or suggesting installation, verify the user's setup:

  1. Ask how they access cuOpt:

    • "Do you have cuOpt installed? If so, which interface?"
    • "What environment are you using? (local GPU, cloud, Docker, server, etc.)"
  2. Different packages by language/interface:

    Language / InterfacePackageCheck
    Pythoncuopt (pip/conda) — also pulls in libcuoptimport cuopt
    Clibcuopt (pip/conda) — already present if cuopt is installedfind libcuopt.so or header check
    REST Servercuopt-server or Dockercurl /cuopt/health
    CLIcuopt package includes CLIcuopt_cli --help

    Note: cuopt declares libcuopt as a runtime dependency, so installing the Python package also installs the C library and headers. Installing libcuopt on its own does not install the Python API.

  3. If not installed, ask how they want to access:

    • "Would you like help installing cuOpt, or do you have access another way?"
    • Options: pip, conda, Docker, cloud instance, existing remote server
  4. Never assume installation is needed — the user may:

    • Already have it installed
    • Be connecting to a remote server
    • Prefer a specific installation method
    • Only need the C library (not Python)
  5. Ask before running any verification commands:

    # Python API check - ask first
    import cuopt
    print(cuopt.__version__)
    
    # C API check - ask first
    find ${CONDA_PREFIX} -name "libcuopt.so"
    
    # Server check - ask first
    curl http://localhost:8000/cuopt/health
    

Ask Before Running

Do not execute commands or code without explicit permission:

ActionRule
Shell commandsShow command, explain what it does, ask "Should I run this?"
Package installsAllowed in user space (pip/conda/Docker) once the user confirms they want cuOpt installed — see below. Only sudo/system-level installs are off-limits.
Examples/scriptsShow the code first, ask "Would you like me to run this?"
File writesExplain what will change, ask before writing

Exceptions (okay without asking):

  • Read-only commands the user explicitly requested
  • Commands the user just provided and asked you to run

No Privileged Operations

🔒 MANDATORY — this is the one non-negotiable refusal. It applies even when the user explicitly asks.

Never do these:

  • Use sudo or run as root
  • Modify system files or configurations (e.g. /etc)
  • Add system-level package repositories or keys
  • Change firewall, network, or driver settings

If a task seems to need one of these, stop and explain what's needed — the user runs the privileged step themselves. Installs into a user-space environment (a virtualenv, a conda env, or the active Python) are not privileged and are covered below.


Installing Packages

Installing cuOpt (and the packages it needs) in user space is allowed — that's what the cuopt-install skill is for. The rule is get the user's go-ahead, not refuse:

  1. Confirm first. Tell the user which package you'll install, which command, and why; install once they agree. (Check the environment first per Check Environment First — they may already have it, or prefer a different method.)
  2. Stay in user space. Use pip, conda/mamba, or Docker into the active env. Never reach for sudo or a system package manager (apt install) — if something seems to need that, surface it and let the user handle the privileged part.
  3. Match the CUDA suffix (-cu12 / -cu13) to the user's runtime, and choose one package manager — don't mix pip and conda for the same package.

Resources

Documentation

Examples

Support