Multi-agent BMAD framework for Claude Code — SPEC → DESIGN → BUILD → QA in one kit ? What would you like to do? › /init-kit — Bootstrap AGENTS.md for a new project (~8 questions) /create-feature — Run the full BMAD pipeline (SPEC → DESIGN → BUILD → QA) /create-spec — BA only — turn a requirement into SPEC.md Quick Start npm i agentcrew-kit npx agentcrew-kit init
No reviews yetBe the first to leave a review for Agentcrew kit AI
Maker
📌
I built agentcrew after a year of using Claude Code on real
client projects and getting tired of re-explaining my stack
every session, watching Claude invent endpoints, and rewriting
specs mid-code.
**agentcrew turns Claude Code into a 12-agent dev team** — BA,
Tech Lead, PM, Designer, Backend/Frontend/Mobile Dev, QA, QC —
each with its own scope. Work flows through a BMAD pipeline
(SPEC → DESIGN → BUILD → QA) with hard phase gates.
Three things I'd highlight:
🔒 **Hooks enforce rules at the OS level** — no committed
secrets, no `git push --force` main, no reading `.env`. Not
"the AI tries not to" — actually blocked.
🔀 **Multi-repo aware** — one spec syncs BE + FE + Mobile via
a locked API contract before parallel work starts.
⚡ **`npx agentcrew-kit init`** and you're running.
MIT licensed, free forever. Default stack is NestJS, React 19,
Flutter + Riverpod — all swappable via `/init-kit`.
Would love to hear: is the BMAD workflow too heavy for solo devs?
What stack should I support next?
→ github.com/longmup/agentcrew-kit