Claude Code Essentials

Claude Code is Anthropic's agentic coding tool. It reads your codebase, edits files, runs commands, and integrates with your development tools — available in the terminal, as IDE extensions (VS Code, Cursor, JetBrains), as a desktop app, and in the browser.

If you'll be using Claude Code for an open-ended interview — or just to build interview-grade habits — here's what you need to know.

Setup

The quickest install on macOS/Linux:

curl -fsSL https://claude.ai/install.sh | bash

Then start it in any project:

cd your-project
claude

You'll be prompted to log in on first use. A paid Claude subscription or Anthropic Console account is required for most surfaces.

Core workflow

Claude Code is designed around a loop: describe what you want → it plans → it edits files across the project → it runs commands to verify → you review the diff and approve.

# Write tests for the auth module, run them, and fix any failures
claude "write tests for the auth module, run them, and fix any failures"

It works directly with git (staging, commit messages, branches, PRs), connects to external systems via the Model Context Protocol (MCP), and can spawn subagents for parallel work.

Plan mode — your rehearsal tool

Claude Code has a plan mode that disables making changes and instead proposes how it'll implement the feature. For interview prep, this is the single most transferable habit: describe the feature, get a plan, iterate on it, and only then switch to executing.

Interview transfer: in a structured interview, there's no plan mode — but the discipline is identical. Form the plan, vet it, then prompt the implementation. Rehearsing with plan mode makes "vet before execute" automatic.

CLAUDE.md and memory

CLAUDE.md is a markdown file you add to your project root that Claude Code reads at the start of every session. Use it to set coding standards, architecture decisions, preferred libraries, and review checklists. Claude also builds auto-memory as it works — saving learnings like build commands and debugging insights across sessions.

<!-- CLAUDE.md -->
- Build command: npm run build
- Test command: npm test
- Lint before every commit
- Keep database access behind the repository layer

Interview transfer: a well-maintained CLAUDE.md is exactly the kind of project intelligence you should absorb during orientation — build/test commands, conventions, and constraints. If your open-ended interview starts from a repo you control, ship it with a CLAUDE.md (or AGENTS.md) so the AI starts with your standards.

Subagents, hooks, and skills

• Subagents — spawn multiple Claude Code agents on different parts of a task simultaneously, with a lead coordinating.
• Hooks — run shell commands before or after actions (auto-format after every edit, run lint before a commit).
• Skills — packaged repeatable workflows your team can share (/review-pr, /deploy-staging).

Interview-relevant habits