Skip to main content
NeVibeCode · repository · 9 items

evals-skills: a set by ai-evals-course

The ai-evals-course/evals-skills repository on GitHub: 9 cards with a description and an install command. They work in Claude Code, Codex, Cursor — 8 platforms in total.

About the repository

GitHub
ai-evals-course/evals-skills
Stars
★ 1,831
Installs across cards
9,894
License
Apache-2.0
Source updated
2026-10-24
In the catalog since
2026-10-25
write-judge-prompt skill write-judge-prompt walks you through one binary Pass/Fail LLM-as-Judge evaluator for a subjective failure mode: tone, faithfulness, relevance. Useful when code checks fall short. CX Codex +1 1,300installs validate-evaluator skill The validate-evaluator skill checks whether an LLM judge agrees with expert labels and measures TPR/TNR on a held-out set. CX Codex +1 1,300installs generate-synthetic-data skill generate-synthetic-data — a skill that builds varied synthetic test queries for LLM pipeline evaluation when real data is sparse. CX Codex +1 1,200installs eval-audit skill eval-audit audits LLM evaluation pipelines to uncover unvalidated judges, vanity metrics, and flawed error analysis. Best used when reviewing existing evals. CX Codex +1 1,200installs error-discovery skill error-discovery — a skill for interactive error analysis on a dataset of LLM outputs: it builds a review UI, picks diverse samples and helps organize failure modes. CX Codex +1 1,200installs build-review-interface skill build-review-interface builds a custom browser-based annotation interface to review LLM traces and record human feedback. Useful when evaluating models and collecting labeled datasets. CX Codex +1 1,100installs evaluate-rag skill evaluate-rag guides separate evaluation of RAG retrieval and generation, metric selection, and document chunking tests. CX Codex +1 1,100installs evals-start skill evals-start routes an evaluation request to the right specialised skill. Use it when you are unsure where to begin or which evaluation workflow fits the task. CX Codex +1 959installs write-code-eval skill write-code-eval is a Claude Code skill for building code evaluators with objective rules. It applies when a known failure can be checked from a trace, with or without a reference answer. CX Codex +1 535installs

More in NeVibeCode

Project sponsors

Companies that keep this analytics open