Skills that guide AI coding agents to help you build product-specific AI evals.
Eval Skills is a collection of instructions for coding agents that helps teams create evaluations for specific AI products, rather than tests of general-purpose foundation models. AI developers and evaluation teams use it to audit pipelines, analyze traces, generate test data, write code-based checks, and assess language-model judges. The catalogue includes nine skills and one plugin from this collection.
Plugin Claude Code
Bundles 9 skills · 547 tokens together
Skills for building LLM evals. Starts with interactive error discovery: build a review app, sample diverse traces, and organize human annotations into failure modes.