Loading video player...
The creator of Claude Code, Boris Cherny, recently shared a detailed thread explaining exactly how he uses Claude Code day to day, from running multiple agents in parallel, to using plan mode, web handoffs, GitHub automation, and shared context via CLAUDE.md. The thread itself is text and screenshots, so in this video I walk through how to actually set up and use the same workflow, step by step, using Claude Code in the terminal, Claude Code Web, and GitHub. If you’re using Claude Code, or planning to, this video is designed as a practical setup guide you can follow along with and adapt to your own projects. ⏱️ TIMESTAMPS 00:00 Intro: The Claude Code Creator’s Workflow 00:45 Running Multiple Claude Code Agents in Parallel (CLI) 01:26 Configuring Notifications in Claude Code 02:16 Avoiding Conflicts & Rolling Back Parallel Agents 02:50 Setting Up Claude Code Web 07:54 Handing Off Tasks from CLI to Claude Code Web 09:05 Teleporting Tasks Back to the CLI 10:24 CLAUDE.md: Shared Context & Workflow Rules 12:04 Creating a Draft CLAUDE.md with Claude Code 12:46 Setting Up the Claude Code GitHub Action 14:29 Using Claude Code During Pull Request Reviews 15:38 Plan Mode vs Accept Edits: When to Use Each 17:08 Wrap-Up & What’s Coming Next 🔗 IMPORTANT LINKS 🧵 Boris Cherny’s original thread on X https://x.com/bcherny/status/2007179832300581177 ⚙️ Claude Code GitHub Action workflow (claude.yml) https://github.com/anthropics/claude-code-action/blob/main/examples/claude.yml 🧠 Learn more about Claude Code: https://claude.com/product/claude-code 💬 HAVE YOU TRIED CLAUDE CODE? How are you using it today — CLI only, Web, or both? Are you running multiple agents in parallel yet? Drop your setup or questions in the comments. 🔔 SUBSCRIBE for real-world AI coding workflows, tooling deep dives, and hands-on tutorials 🌐 Resources & templates: https://snapperai.io 🐦 AI updates on X: https://x.com/SnapperAI