Last commit on August 25, 2026·Created on July 2, 2026
dzhng/skills
“A framework for authoring, testing, and refining agent skills through structured evaluation.”
Combined rank
#334
across all skills
Stars
860
7d change unavailable
Forks
57
Watchers
4
Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of real attention, adoption, and active maintenance.
TL;DR
This collection provides a methodology for creating and hardening agent skills. It enables developers to move from initial skill drafting to rigorous testing against golden cases, using the gaps between actual and expected outcomes to drive iterative improvements.
The toolkit helps maintain a clear distinction between skill logic, reference materials, and scripts, ensuring that agent triggers remain precise and outcomes are predictable.
WHO IT'S FOR
AI agent / automation builders
authoring and refining agent skill sets
Technical leads / architects
auditing AI-driven architectural decisions
Backend / infrastructure engineers
diagnosing AI-assisted performance bottlenecks
Platform / DevEx teams
standardizing AI coding rules and workflows
Want to install it?
Claude Plugin · Clone to Claude · Manual Copy · NPX Skills AddRepository instructions