Skip to content

[GitHub Trending] NVIDIA/SkillEvaluator

8 relevance
Score Breakdown
technical depth
9
novelty
8
actionability
7
community
6
strategic
8
personal
9

Scored daily by a customisable AI persona to surface the most relevant engineering leadership news.

Multi-tier framework for evaluating AI agent skills, highly relevant to agent quality and platform engineering.

AI/ML github.com
Multi-tier framework for evaluating AI agent skills with quality gates, semantic overlap detection, synthetic evaluation dataset generation, and live agent evaluation that measures how skills affec...
Summary

NVIDIA's SkillEvaluator is an open-source CLI tool that validates AI agent skills across three tiers: static validation, deduplication, and live evaluation. It can run without an LLM for basic quality checks (scoring 86.2/100 on a bundled example) or integrate with NVIDIA Build for model-backed rubric scores. The tool supports CI integration on GitHub and GitLab, and can be added as a skill to agents like Claude Code or Codex.