Claude Code

Anthropic's agentic coding tool combining autonomous planning, execution, and validation with extensibility through skills, plugins, MCP servers, and hooks.

Core Concepts

Subagents: Specialized agents (planner, code-reviewer, tester, debugger, docs-manager, ui-ux-designer, database-admin)

Agent Skills: Modular capabilities with SKILL.md + bundled resources (scripts, references, assets) loaded progressively

Slash Commands: User-defined operations in .claude/commands/ expanding to prompts

Hooks: Event-driven shell commands (SessionStart, PreToolUse, PostToolUse, Stop, SubagentStop)

MCP Servers: Model Context Protocol integrations for external tools (GitHub, Jira, databases)

Plugins: Packaged collections distributed via marketplace

Reference Guide

Load references as needed for specific topics:

| Topic | Reference File | Contents | |-------|----------------|----------| | Installation & setup | references/getting-started.md | Prerequisites, installation methods, authentication | | Slash commands | references/slash-commands.md | Full catalog: /cook, /plan, /fix, /test, /docs, /git, /design | | Workflow examples | references/common-workflows.md | Feature implementation, bug fixing, testing, git ops | | Creating skills | references/agent-skills.md | Skill structure, metadata, bundled resources | | MCP servers | references/mcp-integration.md | Configuration, common servers, authentication | | Hooks system | references/hooks-comprehensive.md | Event types, command/prompt hooks, tool matchers | | Plugins | references/hooks-and-plugins.md | Plugin structure, marketplace, installation | | Configuration | references/configuration.md | Settings hierarchy, model config, output styles | | Enterprise | references/enterprise-features.md | SSO, RBAC, sandboxing, audit logging, deployment | | IDE integration | references/ide-integration.md | VS Code extension, JetBrains plugin | | CI/CD | references/cicd-integration.md | GitHub Actions, GitLab workflows | | Advanced features | references/advanced-features.md | Extended thinking, caching, checkpointing | | Troubleshooting | references/troubleshooting.md | Auth failures, MCP issues, performance, debug mode | | API reference | references/api-reference.md | Admin, Messages, Files, Models, Skills APIs | | Best practices | references/best-practices.md | Project organization, security, performance, cost |

Instructions

When answering questions:

1. Identify topic from user query
2. Load relevant reference files (use table above)
3. Provide specific guidance with examples
4. For complex queries, load multiple references

Documentation sources:

• Context7 llms.txt: https://context7.com/websites/claude_en_claude-code/llms.txt?tokens=10000
• Topic search: https://context7.com/websites/claude_en_claude-code/llms.txt?topic=<topic>&tokens=5000
• Official docs: https://docs.claude.com/en/docs/claude-code/
• GitHub: https://github.com/anthropics/claude-code
• Support: support.claude.com