Descriptions:
Greg Isenberg sits down with Remy — known as “AI with Remy” — for a deep dive into what they call “skillsmaxxing”: the practice of packaging repeatable workflows into structured markdown SOPs that Claude Code (and other AI coding agents) can execute consistently without re-explanation.
Remy walks through the full lifecycle of building and distributing skills across a team. The core insight is treating AI agent skills like an app store: individual skill files live in a GitHub repository, a small JSON configuration turns that repo into a plugin, and team members install it as a marketplace source inside Claude Code or Codex. Auto-update is enabled so that when one team member refines a skill, everyone picks up the change automatically. Remy demonstrates department-level segmentation — a copywriter installs the newsletter plugin, not the finance one — and shows how multiple skills can be chained together for complex, multi-step outputs.
The episode covers practical prompts for getting Claude to scaffold the plugin infrastructure itself, how skill chains differ from one-off prompting, and why tight skills eliminate the “slop” that plagues teams relying on raw chat sessions. Remy confirms the approach works across harnesses including Codex, making the framework model-agnostic. Builders looking to systematize AI-assisted operations at the team level will find concrete, copy-paste-ready patterns here.
📺 Source: Greg Isenberg · Published August 19, 2026
🏷️ Format: Podcast
