Claude Code Field Guide

About this knowledge base

How it was built, how to read the status badges, and how to keep adding to it.

This knowledge base started as a single source: a YouTube walkthrough covering 27 Claude Code concepts in about half an hour. Every concept from that video was pulled out, checked against Anthropic's official documentation (code.claude.com/docs) as of September 2026, and marked with a status:

Still accurate

The concept as described in the source video is still accurate today, with no meaningful change.

Updated since the video

The core idea still holds, but the feature has grown, been renamed, or gained new capability since the video was made.

Needs correction

Something in the video is now actively misleading — a changed default, a renamed command, or a feature whose real-world status (e.g. experimental, opt-in) differs from how it was presented.

  • Foundations (01) — What Claude Code actually is, and the basic vocabulary you need before anything else in this library makes sense.
  • Permissions & Safety (02) — How Claude Code decides what it's allowed to do on your machine — and how you dial that up or down.
  • Making It Personal (03) — Giving Claude Code your preferences, project rules and working style so it stops guessing and starts matching how you work.
  • Power Features (04) — The features that turn Claude Code from a conversational assistant into a real automation and extension platform.
  • Parallel & Multi-Agent Work (05) — Running several Claude Code sessions or agents at the same time, without them overwriting each other's work.
  • Automation & Real-World Use (06) — Version control, unattended runs, scheduling, cost, and where Claude Code fits into a real workflow or business.

Eight topics beyond the original 27 were added after research turned up concepts that either didn't exist yet when the video was recorded (agent teams' experimental flag aside, think: sandboxing, plugins, cross-session messaging, CI integrations, the Agent SDK, and working across desktop/web/mobile) or that meaningfully change how you'd use an existing concept today.

This is meant to be added to, not treated as finished. The whole site is plain, dependency-free HTML generated from one Python data file (data.py) and one build script (build.py) — there's no framework or build step required to deploy it, and no framework required to extend it either.

To add a new topic: open data.py, copy an existing entry in the TOPICS list as a template, give it the next sequential num, a unique slug, assign it to one of the existing category values (or add a new category to the CATEGORIES list first), fill in the fields, and run python3 build.py. It regenerates the whole site/ folder, including navigation and prev/next links, automatically.

To revise an existing topic when Claude Code changes again, edit its entry directly — update the status, rewrite status_note to explain what changed, and adjust the body. Re-run the build script and redeploy.