claude-code-best-practice

from vibe coding to agentic engineering - practice makes claude perfect

= Agents · = Commands · = Skills

🧠 CONCEPTS

| Feature | Location | Description | |---------|----------|-------------| | Subagents | .claude/agents/<name>.md | Autonomous actor in fresh isolated context — custom tools, permissions, model, memory, and persistent identity | | Commands | .claude/commands/<name>.md | Knowledge injected into existing context — simple user-invoked prompt templates for workflow orchestration | | | | Knowledge injected into existing context — configurable, preloadable, auto-discoverable, with context forking and progressive disclosure · | | | | | | | | User-defined handlers (scripts, HTTP, prompts, agents) that run outside the agentic loop on specific events · | | | , | Model Context Protocol connections to external tools, databases, and APIs | | | distributable packages | Bundles of skills, subagents, hooks, MCP servers, and LSP servers · · | | | | Hierarchical configuration system · · · · · · | | | | Customizable status bar showing context usage, model, cost, and session info | | | , , , | Persistent context via CLAUDE.md files and imports · · | | | automatic (git-based) | Automatic tracking of file edits with rewind ( or ) and targeted summarization | | | | Command-line flags, subcommands, and environment variables for launching Claude Code · · | | | | Agentic Engineering · Context Engineering · Vibe Coding | | | | Official best practices · · |