Claude Code / core concepts
Extend Claude Code
Integrate Claude Code into team processes through commands, memories, MCPs, Hooks, skills, and SDKs.
Extension level
Extending Claude Code isn't just about MCP. A complete set of extensions usually includes project memory, permission settings, Hooks, Skills, Subagents, Plugins and external tool connections.
| Extension mode | what problem to solve |
|---|---|
| CLAUDE.md / Memory | Inject project context, commands, conventions, and preferences into context. |
| Settings / Permissions | Control default mode, environment variables, tool permissions, and hosting policies. |
| Hooks | Automatically run checking, formatting, notification, or blocking logic before and after tool calls. |
| Skills | Encapsulate repetitive tasks into callable workflows. |
| Subagents | Isolate tasks such as investigation, review, migration, etc. into independent contexts. |
| MCP / Plugins | Integrate external systems and distributable capabilities into Claude Code. |
How to choose
Use the lightest extension method first. If it can be written into CLAUDE.md, don't make it into MCP; if it can be blocked with Hook, don't rely on verbal reminders; repeat the process three or more times, and then turn it into Skill or Plugin.
Project rules
Put CLAUDE.md or .claude/rules, suitable for team sharing.
automatic check
Use Hooks to run formatting, testing, blocking of sensitive commands, or notifications.
external data
Use MCP to connect issues, documents, database read-only queries, or internal tools.
Reusability
Package and distribute using Skills or Plugins.
Extensions also consume context
Each extension may increase context, permissions, or maintenance costs. Especially large MCP schema, long CLAUDE.md, and noisy Hooks output will make the model slower, more expensive, and more difficult to focus.
extension principle
The default is to keep it lightweight and load it according to the needs of the task. Keep long-term rules short and put references into skills or document links so Claude can open them when he needs them.
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.
Expansion capability map
Extending Claude Code requires choosing tools based on your needs. CLAUDE.md solves project rules, settings solves default behavior, permissions control tool boundaries, Hooks automatically execute checks, MCP connects to external systems, Skills solidifies the process, Subagents isolates investigations, Plugins distribution capabilities, and Agent SDK embeds applications.
It is recommended from light to heavy: first write CLAUDE.md; add path rules; then configure permissions; then use Hooks to do mechanical checks; repeat the process to accumulate skills; connect MCP when external data is really needed; use SDK when productization is needed.
More extensions are not always better. Each extension adds context, permissions, and maintenance costs. Overly long rules, noisy hook output, and overly broad MCP permissions can reduce stability.
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 "Extend 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.