shanraisshan / claude-code-best-practice

claude-code-best-practice practice makes claude perfect -white?style=flat&labelColor=555) *Click on this badge to show the latest best practice* *Click on this badge to show implementation in this repo* *Click on this badge to see the Command → Agent → Skill orchestration workflow* Boris Cherny on X ( tweet 1 · tweet 2 · tweet 3 ) 🧠 CONCEPTS | Feature | Location | Description | |---------|----------|-------------| | **Commands** | | Knowledge injected into existing context — simple user-invoked prompt templates for workflow orchestration | | **Subagents** | | Autonomous actor in fresh isolated context — custom tools, permissions, model, memory, and persistent identity | | **Skills** | | Knowledge injected into existing context — configurable, preloadable, auto-discoverable, with context forking and progressive disclosure · Official Skills | | **Workflows** | | | | **Hooks** | | User-defined handlers (scripts, HTTP, prompts, agents) that run outside the agentic loop on specific events · Guide | | **MCP Servers** | , | Model Context Protocol connections to external tools, databases, and APIs | | **Plugins** | distributable packages | Bundles of skills, subagents, hooks, and MCP servers · Marketplaces | | **Settings** | | Hierarchical configuration system · Permissions · Model Config · Output Styles · Sandboxing · Keybindings · Fast Mode | | **Status Line** | | Customizable status bar showing context usage, model, cost, and session info | | **Memory** | , , , | Persistent context via CLAUDE.md files and imports · Auto Memory · Rules | | **Checkpointing** | automatic (git-based) | Automatic tracking of file edits with rewind ( or ) and targeted summarization | | **CLI Startup Flags** | | Command-line flags, subcommands, and environment variables for launching Claude Code · Interactive Mode | | **AI Terms** | | Agentic Engineering · Context Engineering · Vibe Coding | | **Best Practices** | | Official best practices · Prompt Engineering · Extend Claude Code | 🔥 Hot | Feature | Location | Description | |---------|----------|-------------| | **/btw** | | Side chain conversations while Claude is working | | **Code Review** | GitHub App (managed) | Multi-agent PR analysis that catches bugs, security vulnerabilities, and regressions · Blog | | **Scheduled Tasks** | , cron tools | Run prompts on a recurring schedule (up to 3 days), set one-time reminders, poll deployments and builds | | **Voice Mode** | | speak to prompt - /voice to activate| | **Simplify & Batch** | , | Built-in skills for code quality and bulk operations — simplify refactors for reuse and efficiency, batch runs commands across files | | **Agent Teams** | built-in (env var) | Multiple agents working in parallel on the same codebase with shared task coordination | | **Remote Control** | , | Continue local sessions from any device — phone, tablet, or browser · Headless Mode | | **Git Worktrees** | built-in | Isolated git branches for parallel development — each agent gets its own working copy | | **Ralph Wiggum Loop** | plugin | Autonomous development loop for long-running tasks — iterates until completion | See orchestration-workflow for implementation details of **Command → Agent → Skill** pattern. ⚙️ DEVELOPMENT WORKFLOWS 🔥 Hot • Cross-Model (Claude Code + Codex) Workflow • RPI • Ralph Wiggum Loop • Garry Tan (CEO of Y Combinator) - gstack · ★ 15k Others • obra/superpowers · ★ 87k • Github Speckit · ★ 77k • get-shit-done (GSD) · ★ 31k • OpenSpec OPSX · ★ 31k • Brian Casel (Creator of Agent OS) - 2026 Workflow · ★ 4k - it's overkill in 2026 • Human Layer RPI - Research Plan Implement · ★ 1.5k • Andrej Karpathy (Founding Member, OpenAI) Workflow • Boris Cherny (Creator of Claude Code) - Feb 2026 Workflow • Peter Steinberger (Creator of OpenClaw) Workflow 💡 TIPS AND TRICKS ■ **Prompting (4)** • challenge Claude — "grill me on these changes and don't make a PR until I pass your test." or "prove to me this works" and have Claude diff between main and your branch • after a mediocre fix — "knowing everything you know now, scrap this and implement the elegant solution" • Claude fixes most bugs by itself — paste the bug, say "fix", don't micromanage how 👶 • say "use subagents" to throw more compute at a problem — offload tasks to keep your main context clean and focused 👶 ■ **Planning/Specs (5)** • always start with plan mode • start with a minimal spec or prompt and ask Claude to interview you using AskUserQuestion tool, then make a new session to execute the spec • always make a phase-wise gated plan, with each phase having multiple tests (unit, automation, integration) • spin up a second Claude to review your plan as a staff engineer, or use cross-model for review • write detailed specs and reduce ambiguity before handing work off — the more specific you are, the better the output ■ **Workflows (12)** • CLAUDE.md should target under 200 lines per file. 60 lines in humanlayer (still not 100% guaranteed). • use multiple CLAUDE.md for monorepos — ancestor + descendant loading • use .claude/rules/ to split large instructions • use commands for your workflows instead of sub-agents • have feature specific sub-agents (extra context) with skills (progressive disclosure) instead of general qa, backend engineer. • memory.md, constitution.md does not guarantee anything • avoid agent dumb zone, do manual /compact at max 50%. Use /clear to reset context mid-session if switching to a new task • vanilla cc is better than any workflows with smaller tasks • use skills in subfolders for monorepos • use /model to select model and reasoning, /context to see context usage, /usage to check plan limits, /extra-usage to configure overflow billing, /config to configure settings • always use thinking mode true (to see reasoning) and Output Style Explanatory (to see detailed output with ★ Insight boxes) in /config for better understanding of Claude's decisions • use ultrathink keyword in prompts for high effort reasoning • /rename impor…