claude/rules/ to organize instructions by topic. By default, every rule file loads at the start of each session. To make a rule load only when Claude works on specific files, add paths frontmatter: yaml --- paths: - "**/*.ts" --- # TypeScript conventions Prefer interfaces over types. This keeps your main CLAUDE.md lean. TypeScript rules load when Claude reads .ts files, Go rules when it reads .go files. Claude never wades through conventions for languages it isn't touching. 32. Use @imports to keep CLAUDE.md lean Reference docs with @docs/git-instructions.md. You can also reference @README.md, @package.json, or even @~/.claude/my-project-instructions.md. Claude reads the file when it needs it. Think of @imports as "here's more context if you need it" without bloating the file Claude reads every session. 33. Allowlist safe commands with /permissions Stop clicking "approve" on npm run lint for the hundredth time. /permissions lets you allowlist trusted commands so you stay in flow. You'll still get prompted for anything not on the list. 34. Use /sandbox when you want Claude to work freely Run /sandbox to enable OS-level isolation. Writes are restricted to your project directory, and network requests are limited to domains you approve. It uses Seatbelt on macOS and bubblewrap on Linux, so restrictions apply to every subprocess Claude spawns. In auto-allow mode, sandboxed commands run without permission prompts, which gives you near-full autonomy with guardrails. For unsupervised work (overnight migrations, experimental refactors), run Claude in a Docker container. Containers give you full isolation, easy rollback, and the confidence to let Claude run for hours. 35. Create custom subagents for recurring tasks Different from using subagents on the fly (#19), custom subagents are pre-configured agents saved in .claude/agents/. For example, a security-reviewer agent with Opus and read-only tools, or a quick-search agent with Haiku for speed. Use /agents to browse and create them. You can set isolation: worktree for agents that need their own file system. 36. Pick the right MCP servers for your stack The MCP servers worth starting with: Playwright for browser testing and UI verification, PostgreSQL/MySQL for direct schema queries, Slack for reading bug reports and thread context, and Figma for design-to-code workflows. Claude Code supports dynamic tool loading, so servers only load their definitions when Claude needs them. For a comprehensive list of what's available, see our guide on the best MCP servers in 2026. 37. Set your output style Run /config and select your preferred style. The built-in options are Explanatory (detailed, step-by-step), Concise (brief, action-focused), and Technical (precise, jargon-friendly). You can also create custom output styles as files in ~/.claude/output-styles/. 38. Use CLAUDE.md for suggestions, hooks for requirements CLAUDE.md is advisory. Claude follows it about 80% of the time. Hooks are deterministic, 100%. If something must happen every time without exception (formatting, linting, security checks), make it a hook. If it's guidance Claude should consider, CLAUDE.md is fine. 39. Auto-format with a PostToolUse hook Every time Claude edits a file, your formatter should run automatically. Add a PostToolUse hook in .claude/settings.json that runs Prettier (or your formatter) on any file after Claude edits or writes it: json { "hooks": { "PostToolUse": [ { "matcher": "Edit|Write", "hooks": [ { "type": "command", "command": "npx prettier --write \\\\"$CLAUDE_FILE_PATH\\\\" 2>/dev/null || true" } ] } ] } } The || true prevents hook failures from blocking Claude. You can chain other tools too. Add npx eslint --fix as a second hook entry. If you have an editor open to the same files, consider turning off format-on-save while Claude is working. Some