DEV.co
MCP Servers · xu-xiang

everything-claude-code-zh

A comprehensive Chinese-translated configuration and toolkit for Claude Code, bundled as a plugin with 13 agents, 56 skills, and 32 commands. Originated from an Anthropic hackathon winner project; includes production-tested patterns for token optimization, memory persistence, continuous learning, and multi-agent orchestration.

Source: GitHub — github.com/xu-xiang/everything-claude-code-zh
1.6k
GitHub stars
271
Forks
JavaScript
Primary language
MIT
License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

FieldValue
Repositoryxu-xiang/everything-claude-code-zh
Ownerxu-xiang
Primary languageJavaScript
LicenseMIT — OSI-approved
Stars1.6k
Forks271
Open issues1
Latest releaseUnknown
Last updated2026-03-05
Sourcehttps://github.com/xu-xiang/everything-claude-code-zh

What everything-claude-code-zh is

MIT-licensed repository providing agents (planner, architect, code-reviewer, security-reviewer, TDD-guide, etc.), skills (frontend/backend patterns, ClickHouse, content generation, continuous learning with instinct-based evaluation), hooks for cross-session context persistence, rules for multiple languages (TypeScript, Python, Go, Java), and MCP configurations. Distributed as Claude Code plugin with Node.js-based cross-platform (Windows/macOS/Linux) scripts; includes package manager auto-detection and OpenCode/Codex CLI support.

Quickstart

Get the everything-claude-code-zh source

Clone the repository and explore it locally.

terminalbash
git clone https://github.com/xu-xiang/everything-claude-code-zh.gitcd everything-claude-code-zh# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Teams adopting Claude Code for enterprise development

Pre-configured agents for planning, architecture, code review, and security scanning reduce setup time and standardize AI-assisted workflows across Chinese-speaking engineering teams.

Multi-language projects requiring consistent coding standards

Modular language-specific rules (TypeScript/Python/Go/Java/C++) and pattern libraries enable uniform code quality checks and architectural guidance across polyglot codebases.

Continuous learning and pattern extraction from development sessions

Built-in skills for session-to-skill conversion, instinct-based evaluation, and iterative context refinement help teams codify ad-hoc solutions into reusable, evaluated knowledge.

Implementation considerations

  • Plugin installation requires either `/plugin marketplace add` + `/plugin install` (automatic rules fetch not supported; manual clone + `install.sh` recommended) or full manual directory copy.
  • Rules are language-modular (common/ + typescript/ + python/ + golang/ subdirs); only install language-specific rules needed to avoid bloat.
  • Token optimization and memory persistence rely on hooks and cross-session file I/O; projects must support persistent file storage outside Claude Code's ephemeral context.
  • Continuous learning v2 and instinct-based evaluation require domain modeling upfront; generic use without customization may not yield actionable patterns.
  • Multi-agent orchestration (PM2 support, `/multi-plan`, `/multi-execute`, etc.) assumes orchestration tooling (PM2, Docker, or workflow engine) is available in target environment.

When to avoid it — and what to weigh

  • Your team is monolingual (non-Chinese) and English documentation is insufficient — While English README exists, the bulk of guides and extended documentation is Chinese. Translation effort required for non-Chinese teams.
  • You need a tightly integrated, off-the-shelf solution without customization — This is a collection of modular, reference configurations. Effective adoption requires reading guides, understanding agent/skill philosophy, and tailoring to your workflow.
  • Your Claude Code version or API is not actively maintained — Repository targets current Claude Code releases (last pushed March 2026) and Anthropic APIs. Older Claude versions or frozen API contracts may experience compatibility drift.
  • Minimal or no documentation tolerance — While comprehensive, setup and best-practice adoption require careful review of shorthand and longform guides. Quick-start is 2 minutes; expert mastery requires deeper study.

License & commercial use

MIT License. Permissive OSI license allowing unrestricted use, modification, and commercial distribution, provided original license and copyright notice are retained.

MIT license permits commercial use without restriction. No attribution requirement beyond retaining the MIT notice in derivative works. Suitable for proprietary enterprise adoption. However, no warranty or liability protection is provided; organizations should review their standard commercial software evaluation process.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

SignalAssessment
MaintenanceActive
DocumentationStrong
License clarityClear
Deployment complexityModerate
DEV.co fitGood
Assessment confidenceHigh
Security considerations

Considerations only; no security audits or penetration tests claimed. (1) Rules and hooks execute within Claude Code context; code injection risk depends on Claude Code's input validation and sandboxing. (2) AgentShield integration (v1.6.0+) can scan for vulnerabilities; runs as optional `/security-scan` command. (3) Rules include a `security-reviewer.md` agent and `security-review/` skill (self-assessment tool; not external audit). (4) Token optimization and context pruning may mask sensitive data if not carefully handled. (5) Multi-agent orchestration (PM2) could expose inter-process communication to host OS; review PM2 security model for your environment. (6) No explicit claim of secrets management, encryption, or audit logging. Treat as development-time configuration toolkit; ensure secrets (API keys, credentials) are externalized before production use.

Alternatives to consider

Cursor IDE + built-in .cursorrules

Cursor offers native, IDE-integrated agent and rules system without plugin dependency. Simpler for single-language projects; less modular for polyglot/multi-agent workflows. English-primary documentation.

Cline (formerly Claude for VS Code) + manual agent/prompt engineering

Open-source VS Code extension; no pre-built agent templates. Greater customization freedom but higher setup cost. Suitable for teams wanting minimal out-of-the-box assumptions.

Python-based agentic frameworks supporting Claude API; more programmatic, less IDE-native. Better for backend/data pipelines; requires code-level integration vs. configuration-driven setup.

Software development agency

Build on everything-claude-code-zh with DEV.co software developers

Install the Everything Claude Code plugin for pre-built agents, skills, and coding standards. Available for Claude Code, OpenCode, and Codex. Clone the repo, run install.sh, and unlock 13 specialized agents and 56 reusable skills for your AI-assisted development workflow.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

Related on DEV.co

Explore the category and the services that help you build with it.

everything-claude-code-zh FAQ

Does this work without Anthropic's Claude Code, or can I use it with other editors?
Repository is primarily built for Claude Code. v1.6.0+ adds Codex CLI support (generates codex.md). v1.3.0+ supports OpenCode (12 agents, 24 commands, 16 skills via OpenCode plugin events). Cursor and VS Code require manual adaptation of prompts/rules. Not tested against ChatGPT or other LLM IDEs.
What's the difference between agents, skills, commands, and rules?
Agents = specialized subagents for delegated tasks (e.g., planner, architect). Skills = reusable workflow/domain knowledge bundles (e.g., TDD, ClickHouse). Commands = user-callable shortcuts (56 total). Rules = static linting/style configuration files (language-modular). See longform guide for deeper explanation.
Can I use this for non-Claude AI tools (e.g., OpenAI, Anthropic Opus in a different IDE)?
Agents and skills are prompts; extractable for other LLMs. Rules are static; language-agnostic. However, the plugin system, hooks, and command syntax are Claude Code / OpenCode-specific. Manual conversion required. No official cross-platform prompt adaptation provided.
Is this production-ready, or is it best-effort/experimental?
Described as 'production-level configurations' evolved over 10+ months of daily use in a real product. However, no formal SLA, warranty, or security audit backing. Treat as a well-tested reference implementation; validate in your environment before mission-critical use.

Software developers & web developers for hire

From first prototype to production, DEV.co delivers software development services around tools like everything-claude-code-zh. Our software development agency staffs experienced software developers and web developers for custom software development, web development, integrations, and ongoing support across mcp servers and beyond.

Get Started with Claude Code Automation in 2 Minutes

Install the Everything Claude Code plugin for pre-built agents, skills, and coding standards. Available for Claude Code, OpenCode, and Codex. Clone the repo, run install.sh, and unlock 13 specialized agents and 56 reusable skills for your AI-assisted development workflow.