mcpbeat

Nat Optimization

nvidia/nat-optimization

Use when configuring or running NeMo Agent Toolkit optimization with `nat optimize`, including Optuna parameter tuning, prompt evolution, optimizer sizing, output interpretation, and optimizer datasets.

7k tokens
context cost
the whole folder, loaded on every use
7
files
instructions only
0
copies elsewhere
how many repositories repackaged it
2551
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository
npx skills add https://github.com/NVIDIA/NeMo-Agent-Toolkit --skill nat-optimization

What comes with it

25 328 bytes besides the instruction
references/choosing-parameters.md
references/complete-config-example.md
references/configuration.md
references/optimizer_example_dataset.json
references/output-and-cli.md
references/overview.md

The instruction itself

4 sections, as written by the author

NeMo Agent Toolkit Optimization

Use this skill when improving workflow quality through nat optimize.

Workflow

  • Fix workflow correctness issues before optimizing.
  • Size the run and explain the chosen n_trials, parallelism, and stopping behavior.
  • Use separate evaluators for separate quality dimensions.
  • Run nat optimize with a generous timeout.
  • Inspect output artifacts before writing tuned values back to workflow YAML.

Guardrail

Do not kill nat optimize mid-run unless the user asks. It writes final artifacts when the study finishes cleanly.

References

  • references/overview.md
  • references/choosing-parameters.md
  • references/configuration.md
  • references/output-and-cli.md
  • references/complete-config-example.md
  • references/optimizer_example_dataset.json

How to use it

Copy the folder

Take nvidia/nat-optimization from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.