Claude Code / quick start
Overview of Claude Code
Claude Code is an agent-based coding tool that runs where you work, helping you understand your code base, edit files, run commands, and deliver results.
What is Claude Code?
Claude Code is Anthropic's agent-based coding tool for software engineering. Instead of just doing code completion, it reads files in the real code base, understands the context, edits the code, runs commands, and leaves the results for developers to review.
The official documentation describes Claude Code as a coding agent that can work in terminals, IDEs, desktops, browsers, and team collaboration tools. It handles codebase navigation, bug fixing, test generation, refactoring, PR preparation, documentation updates, and ongoing maintenance.
The key for learners is not to use it as a "question and answer robot", but to master a verifiable closed loop: propose a clear task, let Claude read the relevant context, confirm the plan, perform the modification, run the verification, and finally review the diff and command output.
- The terminal is suitable for real repository commands and local verification.
- The IDE is great for viewing context, referencing files, and reviewing inline diffs.
- Portals such as Web, Desktop, Slack, and GitHub are suitable for parallel tasks and team collaboration.
core working methods
Claude Code enters an agent-like loop around the goal: gather context, plan, use tools, read results, continue tweaking until the task is complete or requires your approval.
Give task
Describe the objectives, scope, limitations, and acceptance criteria. Don't just say "optimize it", make it clear where, why, and how to accept it.
read code
Claude reads files, searches for symbols, and views tests and configurations. You can ask it to explain the findings first, without revising them right away.
Execution and verification
Allows it to edit files, run commands, view failure output, and iterate to passing validation.
Review delivery
Finally look at the diff, command output, remaining risks, and next steps rather than just one sentence complete.
Choose to use the entrance
The capability boundaries of different entrances are different. Newbies are advised to start with the terminal or VS Code and become familiar with permission mode, context window, CLAUDE.md and common workflows before using parallel portals such as Web, Desktop, and Slack.
| entrance | suitable for what | Things to note |
|---|---|---|
| CLI | Local repository, command verification, scripted operation. | The permission mode and shell environment determine what it can do. |
| VS Code / JetBrains | Reference current file, inline diff, in-editor review. | Still have to run real builds and tests. |
| Web / Desktop | Cloud or multi-session tasks, cross-device relay, long task management. | Local login status and private tools may not be available. |
| Slack / GitHub | Initiate tasks from team conversations, issues, PRs. | The final merge still comes back to code review and CI. |
Complete usage points
Supplement the core concepts, operation sequences, permission boundaries and verification requirements that are easily compressed and missed in official documents, making it easier for English readers to learn completely by page.
Official document main line
The main line of the official Claude Code documentation is "Let Claude work directly on the code base where you work." It covers CLI, IDE, Web, Desktop, Chrome, Slack, GitHub, MCP, Hooks, Skills, Subagents, Settings, Sessions and Agent SDK. When getting started, you should first master the terminal workflow, and then gradually expand to the team and automation portal.
Suggested order of study: overview and quick start, how it works, CLAUDE.md and memory, context window, permission mode, common workflows, platform integration, MCP/Hooks/Settings, Agent SDK.
The core of Claude Code is not to "generate code" but to perform tasks around a real repository. You need it to read context, propose plans, edit files, run commands, validate results, and retain human approval before critical operations.
Study Checklist
Put the content on this page into real tasks and use the five dimensions of entry, context, permissions, verification and team rules to check whether you have truly mastered it.
Study Checklist
After reading this page, do not just remember the concept name. You should be able to place "Overview of Claude Code" back into a real Claude Code engineering workflow: where the task starts, what context the system loads, which actions need approval, how the result is verified, and how to roll back when it fails.
If this is a concept page, be specific about how it affects the real task: does it change context, permissions, execution paths, validation methods, or changes the team collaboration process.
- Be able to describe in your own words the specific problem this page solves, rather than just reciting the title.
- Able to write a minimal example task with goals, scope, prohibitions, and acceptance criteria.
- Be able to determine which information should be put into the current prompt and which should be captured as project rules or configurations.
- Be able to explain which long-term rules should go into CLAUDE.md, and which runtime behavior should be handled by settings, permissions, Hooks, Skills and MCP.
- Ability to check diffs, command output, test results, screenshots or PR notes after a task is completed instead of just trusting the natural language summary.
If this page is used for team training, ask learners to complete a small task with Claude Code: read and explain first, submit a plan, make the smallest useful change, and close with real verification commands plus human diff review.