Productivity & WorkflowšŸ¤–claudešŸ’¬gptāŒØļøcursor🌐universal

Claude Configuration Management System

Self-use Claude Code rules / settings / permissions / commands, use claude command to sync rules for Droid/OpenCode/Codex/Qwen/Cursor/Copilot. Not fully tested, use at your own risk.

Quick Start

Install this Skill

npx skills add CsHeng/dot-claude
# Claude Configuration Management System

A comprehensive configuration management and agent orchestration system for Claude Code environments. This repository provides a unified framework for synchronizing rules, agents, skills, and commands across multiple AI CLI targets.

## Overview

This system enables centralized management of Claude Code configurations with support for multiple target environments including Droid CLI, Qwen CLI, OpenAI Codex CLI, OpenCode, and Amp CLI. It provides automated synchronization, backup management, and governance capabilities.

### Architecture: User-Level vs Project-Level

The system operates across two complementary levels:

- **User-Level** (`~/.claude/`): Global configuration and personal automation tools that apply to all projects
- **Project-Level** (`.claude/` within projects): Project-specific management tools (config-sync) scoped to individual projects

When Claude Code runs in the `~/.claude/` directory, it merges both levels for development purposes. In normal projects, only user-level components are available.

### Discovery Model

Agents, skills, and commands are discovered automatically via their frontmatter - no manual registration required:
- **Commands** (`commands/*.md`): User-visible slash commands
- **Agents** (`agents/*/AGENT.md`): Execution units with system prompts
- **Skills** (`skills/*/SKILL.md`): Reusable capability modules
- **Rules** (`rules/*.md`): Development standards (auto-loaded by CLAUDE.md)

## Key Components

### šŸ”§ Configuration Synchronization (`config-sync`)
- **Multi-target support**: Synchronize configurations across different AI CLI environments
- **Automated backup**: Built-in backup and retention policies
- **Phase-based execution**: Structured workflow with collect → analyze → plan → prepare → adapt → execute → verify → cleanup → report
- **Target adapters**: Specialized adapters for each CLI environment
- **Project-level component**: Located at `~/.claude/.claude/` for project-specific management

### šŸ¤– Agent System
Specialized agents for different workflows:

**User-Level Agents** (available globally):
- `agent:llm-governance`: LLM prompt optimization and governance
- `agent:workflow-helper`: Draft commit messages and shell script review
- `agent:code-architecture-reviewer`: Architecture review and compliance
- `agent:code-refactor-master`: Code refactoring and restructuring
- `agent:plan-reviewer`: Development plan review and validation
- `agent:ts-code-error-resolver`: TypeScript error resolution
- `agent:web-research-specialist`: Research and information gathering
- `agent:refactor-planner`: Complex refactoring planning

**Project-Level Agents** (project-specific management):
- `agent:config-sync`: Configuration synchronization and management

### šŸ› ļø Skills Framework
Domain-specific skills providing focused expertise:
- **Language skills**: Python, Go, Shell scripting standards
- **Architecture skills**: Patterns, development standards, security
- **Workflow skills**: Discipline, automation selection, environment validation
- **Governance skills**: LLM governance, output style management
- **Quality skills**: Testing strategy, error patterns, quality standards

### šŸ“‹ Rule System
Comprehensive rule set covering:
- Development standards and best practices
- Security standards and guardrails
- Communication protocols and output styles
- LLM prompt writing guidelines
- Language-specific guidelines (Python, Shell, Go)
- Cross-language architecture principles

## Directory Structure

### User-Level Structure (`~/.claude/`)
Global configuration available across all projects:
```
~/.claude/
ā”œā”€ā”€ CLAUDE.md                 # Rule-loading conditions
ā”œā”€ā”€ AGENTS.md                 # Agent discovery documentation
ā”œā”€ā”€ rules/                    # Development standards (auto-loaded)
ā”œā”€ā”€ skills/                   # User-level skill definitions
ā”œā”€ā”€ agents/                   # User-level agent definitions
ā”œā”€ā”€ commands/                 # User-level command definitions
ā”œā”€ā”€ output-styles/            # Named output style manifests
ā”œā”€ā”€ docs/                     # Documentation
ā”œā”€ā”€ settings.json             # Global configuration
└── README.md                 # This file
```

### Project-Level Structure (`.claude/` within projects)
Project-specific Claude Code management tools:
```
.claude/
ā”œā”€ā”€ CLAUDE.md                 # Project-level rule overrides (inherits user-level)
ā”œā”€ā”€ skills/                   # Project-specific skills
ā”œā”€ā”€ agents/                   # Project-specific agents
ā”œā”€ā”€ commands/                 # Project-level commands
ā”œā”€ā”€ config-sync/              # Config-sync subsystem
│   ā”œā”€ā”€ sync-cli.sh          # Unified orchestrator
│   ā”œā”€ā”€ settings.json        # Sync configuration
│   ā”œā”€ā”€ adapters/            # Target-specific adapters
│   ā”œā”€ā”€ lib/                 # Shared libraries and phases
│   └── scripts/             # Utility scripts
```

## Quick Start

### Prerequisites
- Claude Code CLI
- Shell environment (bash/zsh)
- Optional: P

... (truncated)
Repository
7stars
0forks
View on GitHub
Details
DiscoveredJan 19, 2026
Last UpdatedJan 18, 2026
Author@CsHeng
Quality Score
40/100