build-with-dhiraj/ai-workflow-framework-portability-kit

🧰 AI Workflow Framework & Portability Kit

Your entire Claude Code stack — 35 agents, 115 skills, plugins, MCP & tooling — portable across Macs. Clone, claude login, run one script, working in ~20 min.

<p> <a href="LICENSE"><img alt="License" src="https://img.shields.io/github/license/build-with-dhiraj/ai-workflow-framework-portability-kit?color=yellow"></a> <img alt="last commit" src="https://img.shields.io/github/last-commit/build-with-dhiraj/ai-workflow-framework-portability-kit"> <img alt="agents" src="https://img.shields.io/badge/agents-35-blue"> <img alt="skills" src="https://img.shields.io/github/directory-file-count/build-with-dhiraj/ai-workflow-framework-portability-kit/Skills?type=dir&label=skills&color=blue"> <img alt="platform" src="https://img.shields.io/badge/macOS-000000?logo=apple&logoColor=white"> <img alt="Claude Code" src="https://img.shields.io/badge/Claude%20Code-orchestration-D97757"> </p>

A portable, self-contained snapshot of a working Claude Code setup — the specialist agents, the skill library, the plugins, the MCP servers, the host-side tooling, and the orchestration logic that ties them all together. Drop the folder on a fresh Mac, run one script, and the entire stack is restored in about twenty minutes.

This is the master config behind a multi-layer "Engineering Manager" AI workflow: a top-level orchestrator that never writes code itself, dispatching work down to 35 specialist agents who do. The kit captures that architecture as plain files anyone can audit, fork, or rebuild from.

Want the deep model? This README is the landing page. CLAUDE.md is the full deep-dive — the layered architecture, every skill-precedence rule, and the agent org chart.

---

🤔 Why this exists

LLM coding agents got powerful enough that the bottleneck stopped being "can the model do it" and started being "does my setup let it do the right thing without me babysitting."

That setup — agents + ski