How do I set up VS Code as a local AI workspace?

VS Code is a free code editor, but in the book it plays a different role: a local command center where your project documents, AI assistant, and scripts sit side by side. Chapter 11, Local AI Environment, covers the setup in detail. You don't need it for Chapters 1–8.

Basic setup

1. Install VS Code from Microsoft's official site. Check your organisation's software policy first if you're on a work machine.
2. Create a project folder, for example website-relaunch, and open it in VS Code.
3. Organise it the way the book organises the Brain, with one subfolder per tier:
   • constitution for charter, scope, stakeholders, and glossary
   • news for meeting notes, decisions, and weekly updates
   • library for detailed plans, specs, and past reports
   • outputs for reports and artifacts the AI generates
4. Save documents as plain text or Markdown where you can. AI tools read these easily, and they're simple to update.
5. Add an AI assistant. The book uses Claude Code. See What is Claude Code, and do I need it?. Follow the provider's current installation guide, since steps change.

Make it a habit

• Keep a context brief at the top of the folder: goal, deadline, key people, current phase, top risks.
• Date your files so the assistant can tell current from old.
• Clean up weekly. Archive superseded files so they don't confuse the AI.
• Back up the folder using whatever your organisation approves, since it's now your project's memory.

What comes next

Once the workspace is running, Chapter 12 adds Python scripts and MCP servers, and Chapter 13 builds a Next.js dashboard on top. See Do project managers really need Python?.

For a gentler introduction, read Local AI Setup Is Too Complicated and Building a Command Center Sounds Too Technical.