Claude Code Configuration Repository

This repository contains your custom Claude Code configurations including settings, commands, agents, and hooks.

📁 Repository Structure

.
├── .claude/
│   ├── settings.json      # Global settings (permissions, hooks, models)
│   ├── commands/          # Custom slash commands
│   │   ├── clean-up-unused.md   # Remove unused code
│   │   ├── docs.md              # Documentation generator
│   │   ├── merge-conflict-resolver.md # Resolve merge conflicts
│   │   ├── pr_fix.md            # Fix PR review comments
│   │   ├── pr_review.md        # Pull/Merge request reviewer
│   │   ├── repo-research.md    # Repository analysis
│   │   ├── review.md           # Code review helper
│   │   └── test-all.md         # Run all tests
│   ├── agents/            # Custom subagents
│   │   ├── architect.md
│   │   ├── code-reviewer.md
│   │   ├── debugger.md
│   │   └── test-writer.md
│   └── skills/            # Custom skills
│       ├── commit/
│       │   └── SKILL.md   # Smart commit message generator
│       └── github-cli/
│           ├── SKILL.md   # Routing + shared rules for `gh` workflows
│           └── references/
│               ├── issue.md  # Standard for opening a GitHub issue
│               └── pr.md     # Standard for opening a GitHub PR
├── .mcp.json              # MCP server configuration
├── .env.example           # Environment variables template
├── config.sh              # Shared configuration (sourced by install/status)
├── install.sh             # Installation script (copies files to targets)
├── status.sh              # Check configuration status
└── README.md              # This file

🚀 Installation

First Time Setup

1. Clone this repository:

   git clone <your-repo-url>
   cd AI-Configs

2. Set up environment variables:

   cp .env.example .env
   # Edit .env with your API tokens and target directories

   Add one or more CONFIG_SOURCE_DIR entries pointing to your Claude data directories:

   CONFIG_SOURCE_DIR=~/.claude-provectus
   CONFIG_SOURCE_DIR=~/.claude-livenation

3. Run the install script:

   ./install.sh

This copies commands, agents, skills, and .mcp.json from this repo into each target directory listed in .env. Stale files that no longer exist in the source are automatically removed.

What Gets Installed

• commands/ → Custom slash commands
• agents/ → Custom subagents
• skills/ → Custom skills
• .mcp.json → MCP server configuration

🔄 Checking Status

To check the configuration status:

./status.sh

📝 Configuration Files

settings.json

Contains:

• Tool permissions
• Model preferences
• Hook configurations
• Feature flags

commands/

Custom slash commands stored as Markdown files.

Example (.claude/commands/review.md):

1. Run git diff to see recent changes
2. Review code for bugs, security issues, and best practices
3. Provide detailed feedback with line numbers
4. Suggest improvements

Usage: /review

agents/

Custom subagents with specialized roles.

Example (.claude/agents/test-writer.md):

---
name: test-writer
description: Writes comprehensive unit tests for code
tools: Read, Write, Edit, Bash
model: sonnet
---

You are a test writing expert. When invoked:
1. Analyze the code structure
2. Write comprehensive unit tests
3. Follow testing best practices
4. Ensure high coverage

🎯 Usage

Available Skills

/commit - Intelligent commit message generator (skill)

• Analyzes your repository's commit history
• Detects and follows existing conventions
• Generates 3 commit message candidates
• Interactive selection process

/github-cli - Project-agnostic GitHub issue + PR opener (skill)

• Single entry point that routes to either the issue or PR workflow
• Detects the default branch, package manager, and CI gates from the repo
• Optionally runs lint / format / typecheck / build / test before opening a PR
• Learns the project's voice from recent issues and merged PRs
• Drafts title and body, confirms with you, then opens via gh issue create / gh pr create

Available Slash Commands

/pr_review [number] - Comprehensive PR/MR review

• Reviews code quality, security, and performance
• Provides categorized feedback (Critical, High, Medium, Low)

/pr_fix [number] - Fix PR/MR review comments

• Fetches all review comments
• Categorizes and prioritizes issues
• Implements fixes automatically
• Creates clean commits with fixes

/review - Quick code review

• Reviews recent git changes
• Provides feedback on quality and best practices
• Suggests improvements

/test-all - Run all tests

• Executes test suite
• Analyzes failures
• Suggests fixes

/clean-up-unused - Remove unused code

• Identifies dead code, unused imports, and unreferenced functions
• Cleans up safely

/merge-conflict-resolver - Resolve merge conflicts

• Analyzes conflicting changes
• Resolves conflicts intelligently

/docs - Generate documentation

• Analyzes codebase structure
• Identifies undocumented code
• Generates comprehensive documentation
• Follows language-specific conventions

/repo-research - Repository analysis

• Analyzes languages, frameworks, and tech stack
• Lists dependencies and their versions
• Maps project structure and architecture
• Identifies build tools, testing, and CI/CD setup
• Flags outdated dependencies and security concerns

Usage Examples

# Generate smart commit message
/commit

# Review PR #42
/pr_review 42

# Fix issues in current branch's PR
/pr_fix

# Quick code review
/review

# Generate documentation
/docs

# Analyze repository structure
/repo-research

Custom Agents

Agents are invoked automatically by Claude when relevant, or explicitly:

Use the test-writer agent to add tests for this function
Use the code-reviewer agent to review this file
Use the debugger agent to find and fix this bug
Use the architect agent to review the system design

Available Agents:

• code-reviewer - Expert code reviewer for quality, security, and best practices
• test-writer - Writes comprehensive unit and integration tests
• debugger - Systematic debugging and troubleshooting expert
• architect - Software architecture and design specialist

🔧 Customization

Adding a New Command

1. Create a new .md file in .claude/commands/:

   touch .claude/commands/my-command.md

2. Add numbered instructions:

   1. First step
   2. Second step
   3. Third step

3. Commit and push:

   git add .claude/commands/my-command.md
   git commit -m "Add my-command"
   git push

Adding a New Agent

1. Create a new .md file in .claude/agents/:

   touch .claude/agents/my-agent.md

2. Add frontmatter and instructions:

   ---
   name: my-agent
   description: What this agent does
   tools: Read, Write, Edit
   model: sonnet
   ---
   
   Your system prompt here...

3. Commit and push

🔗 MCP (Model Context Protocol) Setup

This repository includes MCP server configurations for Sequential Thinking and Context7.

Sequential Thinking:

• No additional configuration required
• Provides extended thinking capabilities for complex reasoning tasks

Context7:

• No additional configuration required
• Provides up-to-date code documentation and examples
• Optional API key available for higher rate limits and private repos

Environment Variables

All sensitive data is stored in .env (not committed to git). See .env.example for all available variables:

• CONFIG_SOURCE_DIR - Target Claude data directory (one or more)

🔗 How Installation Works

This setup uses copy-based installation. Running ./install.sh copies managed items from this repo into each target directory configured in .env. Running it again overwrites targets with the latest source and removes stale files.

Shared configuration (colors, paths, managed item lists, .env parsing) lives in config.sh and is sourced by both install.sh and status.sh. To manage a new folder or file, add it to the COPY_DIRS or COPY_FILES array in config.sh.

📚 Resources

• Claude Code Documentation
• Custom Commands Guide
• Subagents Guide
• Hooks Guide

🤝 Contributing

If you're sharing this with a team:

1. Fork this repository
2. Make your changes
3. Submit a pull request
4. Run ./install.sh to deploy latest changes

📄 License

Your choice of license here.