claude-code-best-practice
from vibe coding to agentic engineering - practice makes claude perfect
claude-code-best-practice is a comprehensive best-practice repository for Claude Code that transforms informal coding into systematic agentic engineering. Created by shanraisshan, it provides organized collections of subagents, commands, skills, and orchestration workflows alongside implementation examples for Claude Code's features. The repository catalogs proven development workflows from the community, compares popular methodologies, and offers 83+ actionable tips on prompting, session management, context optimization, and Git workflows to help developers build production-ready AI engineering systems.
Key Features
Use Cases
- 01Setting up structured Claude Code projects with organized subagents, commands, and skills following proven patterns
- 02Implementing orchestration workflows that coordinate multiple agents and skills for complex multi-step tasks
- 03Optimizing context window usage and session management to avoid degraded results in long-running tasks
- 04Building cross-model development workflows that combine Claude Code with other AI assistants for specialized review or execution
- 05Learning community-tested techniques for prompt engineering, planning, and code review with Claude Code
- 06Configuring production-ready CLAUDE.md and .claude/rules files that ensure consistent agent behavior across teams
Related Agents
View moreclaude-code-best-practice — FAQ
What is claude-code-best-practice?+
claude-code-best-practice is a GitHub repository that curates best practices, patterns, and implementations for building AI engineering systems with Claude Code. It includes working examples of subagents, commands, skills, orchestration workflows, and a comprehensive collection of community tips for effective Claude Code usage.
How do I install or use claude-code-best-practice?+
Clone the repository from github.com/shanraisshan/claude-code-best-practice and explore the .claude/ directory for example agents, commands, and skills. Copy patterns and configurations into your own Claude Code projects, or reference the best-practice guides and implementation examples to build your own workflows.
Which AI clients work with the patterns in this repository?+
The repository focuses on Claude Code and Claude Desktop. It also documents cross-model workflows for integrating with Codex, Gemini, GPT, Cursor, and other AI coding assistants through plugins, MCP servers, or API routing.
Do I need API keys or prerequisites to use these best practices?+
You need Claude Code installed and configured with appropriate API access. Some workflows reference external MCP servers, plugins, or cross-model integrations that may require additional API keys (OpenAI, Google, etc.), but the core best practices and patterns work with Claude Code alone.
Is claude-code-best-practice free to use?+
Yes, the repository is open source and free to use. However, using Claude Code itself requires a subscription or API access to Anthropic's Claude models. Third-party integrations mentioned (Codex, Gemini, etc.) have their own pricing.
What is the orchestration workflow pattern?+
The orchestration workflow pattern is a Command → Agent → Skill architecture demonstrated in the repository where a slash command invokes a specialized agent that uses one or more skills to complete a task. This pattern enables progressive disclosure of capabilities and keeps context clean by delegating work to focused subagents.
How do I install claude-code-best-practice?+
Open the source repository on GitHub and follow its README. claude-code-best-practice is a agent — MCP Agents Market links you directly to the official repo.
Is claude-code-best-practice free?+
claude-code-best-practice is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.