Implementing CLAUDE.md and Agent Skills In Your Repository
A practical guide to the 3-tier documentation architecture that makes AI coding agents work: root CLAUDE.md, task-specific skills, and agent guides with progressive disclosure.
I'm Matthew Groff, Principal AI Engineer and AI Capability Lead at Umbrage, part of Bain & Company. Based in Orlando, FL, I ship AI agents and MCP servers that automate business operations with AI-enabled workflows. I turn AI concepts into measurable outcomes, cutting processes from months to minutes. I recently published my first eBook “Build Your First MCP Server” and share insights on AI engineering and web technologies through this development blog.
Office Hours: I hold monthly sessions to help teams tackle AI engineering challenges. Book a 30-minute session to discuss your AI Product, MCP Servers, Web Dev, Systematically improving your app with Evals, or whatever strikes your fancy. These times are odd because it's weekends and before and after my day job.
Professional Disclaimer: All content and opinions expressed here are my personal work and views. They are not affiliated with, endorsed by, or representative of Umbrage or Bain & Company. Groff Dev LLC is a separate legal entity that operates independently from my employment obligations.
A practical guide to the 3-tier documentation architecture that makes AI coding agents work: root CLAUDE.md, task-specific skills, and agent guides with progressive disclosure.
A practical update on my early 2026 workflow with OpenCode and Claude Code, including Ralph loops, human gatekeeping, and reusable agent skills.
OpenCode and other AI tools are requesting text/markdown by default. Here's why you should support Accept headers for better AI agent compatibility.
My experience switching from Cursor to Claude Code Max and OpenCode, and how it transformed my development workflow while building an AI Agent Platform.
