NeMo Agent Toolkit Evaluation

SkillDatabases & data

Lets your agent design, run, and troubleshoot evaluations that measure how well your NeMo Agent Toolkit agents perform.

Then ask your AI: use the NeMo Agent Toolkit Evaluation skill

Details

Instructions available. Your AI can read the instructions. Execution depends on the setup they require.

Add it as a step in a pipeline you build on Ahel, or connect Ahel to Claude, ChatGPT, Cursor, Claude Code or Codex and ask your AI to use it.

About this skill

Use when designing, configuring, running, or troubleshooting NeMo Agent Toolkit evaluations, datasets, evaluator selection, ATIF surfaces, quality gates, custom evaluators, and `nat eval`.

What this skill tells your AI

The instructions your AI receives, as published by nvidia/nemo-agent-toolkit in skills/nat-evaluation/SKILL.md and read by Ahel’s review.

Use this skill for measuring agent quality and behavior.

Workflow

  1. Decide the evaluation surface and output format.
  2. Decompose quality goals into separate evaluators.
  3. Choose built-in evaluators before writing custom evaluators.
  4. Keep datasets small and explicit for local validation.
  5. Run nat eval and inspect generated artifacts.

References

  • references/operating-mode.md
  • references/methodology.md
  • references/agent-eval-framework.md
  • references/evaluation-surfaces.md
  • references/evaluation-contract.md
  • references/evaluators/
  • references/code-patterns.md

Signals

GitHub stars
3k
Forks
780
Last commit
Oct 2026

Ahel review

  • K1binfo
    installs-packages (in references/code-patterns.md)
  • K1binfo
    installs-packages (in references/evaluators/evaluator-langsmith-judge.md)
  • K1binfo
    installs-packages (in references/evaluators/evaluator-langsmith.md)

Automated review, not a security audit. Ruleset v1+k2.

Advanced
Item type
skill
Key
nat-evaluation
Source
github.com/nvidia/nemo-agent-toolkit