MoAI-ADK — Spec-First Claude Code Agent Kit

Main

MoAI-ADK is opinionated infrastructure. The value is not only the skill count, but the fact that roles, hooks, and quality gates are bundled into one coherent operating model.

Its published numbers matter: 24 agents, 52 skills, 18 languages, and 27 hook events signal that the maintainers are optimizing for breadth and consistency rather than a tiny demo.

Evaluate it when you want a strong default system for Claude Code adoption across a team and you are willing to accept its spec-first philosophy.

Source-backed notes

• README states 24 specialized AI agents and 52 skills.
• The Go rewrite claims 18 programming languages and 27 Claude Code hook events.
• Key numbers also include roughly 38,700+ lines of Go and 85-100% test coverage.

FAQ

Q: What makes it different from a loose prompt pack? A: It bundles agents, skills, quality gates, and hook events into one environment.

Q: Is it lightweight? A: It is distributed as a Go-based single-binary environment rather than a large Python dependency chain.

Q: Who should try it? A: Teams standardizing Claude Code usage across many repos or languages.