Claude Recommended Skills
The 9 skill categories Anthropic's own team uses, as installable Claude Code skills.
About
VersoXBT/claude-recommended-skills implements the nine high-impact Claude Code skill categories described by Anthropic engineer Thariq in 'Lessons from Building Claude Code': library-api-reference, product-verification, code-scaffolding, code-quality-review, cicd-deployment, data-fetch-analysis, business-process-automation, runbook-investigator, and infra-operations.\n\nEach skill ships a lean SKILL.md with trigger-focused descriptions, 'gotchas' from production use, progressive-disclosure reference docs, and executable scripts (e.g. pipeline health checkers). It is MIT-licensed and installs via a one-liner into ~/.claude/skills. Limitation: a small community repo (single-digit stars at last check) implementing one engineer's writeup rather than an Anthropic-maintained package — treat the patterns as guidance, not gospel.
What it can do
- 9 production skill patterns (review, CI/CD, runbooks, infra ops)\nProgressive-disclosure reference documents\nExecutable scripts for pipeline and infra checks\nOODA-loop incident investigation templates
Setup
npx skills add https://github.com/VersoXBT/claude-recommended-skills or copy skill folders into ~/.claude/skills.
Prerequisites
Claude Code; Git.
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 61; Trending score 70/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.