Claude Code Config

Pre-configured Claude Code setup with specialized agents and workflow automation.

Requirements

uv - Fast Python package manager (used for running hook scripts)

Installation

Option 1: Clone directly to ~/.claude

⚠️ If ~/.claude already exists, back it up first! See backup instructions below.

git clone https://github.com/myk-org/claude-code-config ~/.claude

See also:

Backup Existing Config - If you have an existing ~/.claude
GNU Stow Integration - For dotfiles users

Option 2: Clone as a regular git repo

Clone to any location (e.g., ~/git/), then set up ~/.claude using one of these methods:

git clone https://github.com/myk-org/claude-code-config ~/git/claude-code-config

Then choose how to use it:

Symlink Approach - Symlink components to ~/.claude
Copy Approach - Copy files to ~/.claude

Backup Existing Config

If you have an existing ~/.claude directory:

# Backup existing config
mv ~/.claude ~/.claude.backup

# Clone the repo
git clone https://github.com/myk-org/claude-code-config ~/.claude

# Copy your private files back (examples - adjust to your setup)
# Private agents:
cp ~/.claude.backup/agents/my-private-agent.md ~/.claude/agents/
# MCP server configs:
cp -r ~/.claude.backup/servers/code-execution/configs/* ~/.claude/servers/code-execution/configs/
# Any other private files you have...

# Remove backup after verifying everything works
rm -rf ~/.claude.backup

Symlink Approach

Clone to a different location and symlink into your existing ~/.claude:

# Clone to ~/git/ (or your preferred location)
git clone https://github.com/myk-org/claude-code-config ~/git/claude-code-config

# Create ~/.claude if it doesn't exist
mkdir -p ~/.claude

# Symlink each component
ln -sf ~/git/claude-code-config/agents ~/.claude/agents
ln -sf ~/git/claude-code-config/commands ~/.claude/commands
ln -sf ~/git/claude-code-config/scripts ~/.claude/scripts
ln -sf ~/git/claude-code-config/servers ~/.claude/servers
ln -sf ~/git/claude-code-config/settings.json ~/.claude/settings.json
ln -sf ~/git/claude-code-config/statusline.sh ~/.claude/statusline.sh

Note: Your private configs in ~/.claude/servers/code-execution/configs/ will be inside the symlinked directory. You may want to keep configs outside the repo.

Copy Approach

Clone to a different location and copy files to your existing ~/.claude:

# Clone to a temp location
git clone https://github.com/myk-org/claude-code-config /tmp/claude-code-config

# Copy to ~/.claude (won't overwrite existing files)
cp -rn /tmp/claude-code-config/* ~/.claude/

# Or force overwrite (careful with private configs!)
# cp -r /tmp/claude-code-config/* ~/.claude/

# Clean up
rm -rf /tmp/claude-code-config

Updating with this approach:

git clone https://github.com/myk-org/claude-code-config /tmp/claude-code-config
cp -r /tmp/claude-code-config/agents ~/.claude/
cp -r /tmp/claude-code-config/scripts ~/.claude/
# ... selectively copy what you need
rm -rf /tmp/claude-code-config

Integration with Dotfiles (GNU Stow)

If you manage your dotfiles with GNU Stow, use this clone + overlay approach.

How It Works

Step 1: git clone this repo directly to ~/.claude (base config)
Step 2: stow your dotfiles (private .claude files overlay on top)

Directory Structure

This repo (cloned directly to ~/.claude):

~/.claude/                    ← git clone destination
├── agents/                   # Public agents (from this repo)
├── commands/                 # Commands
├── scripts/                  # Public scripts
├── servers/                  # Server configs
├── settings.json             # Base settings
└── statusline.sh

Your dotfiles (private overlay via stow):

dotfiles/
├── .zshrc
├── .config/
└── .claude/                  # Only YOUR private additions
    ├── agents/               # Private agents
    │   ├── my-private-agent.md
    │   └── company-specific-agent.md
    ├── scripts/              # Private scripts
    │   └── my-helper.sh
    ├── commands/             # Private commands
    │   └── my-workflow.md
    └── servers/
        └── code-execution/
            └── configs/      # Your MCP server configs
                ├── my-server.json
                └── company-internal-service.json

Setup

Note: If ~/.claude already exists, see Option 1 for backup instructions before cloning.

# 1. Clone this repo to ~/.claude
git clone https://github.com/myk-org/claude-code-config ~/.claude

# 2. Stow your dotfiles (overlays private files)
cd ~/dotfiles && stow -t ~ .

Updating

# Update base config:
cd ~/.claude && git pull

# Your private files from dotfiles remain untouched

What's Included

23 specialized agents for different domains (Python, Go, Java, Docker, Kubernetes, Git, etc.)
Orchestrator pattern with automatic agent routing via CLAUDE.md
Pre-commit hooks for rule enforcement
Status line integration

Agents

Agent	Purpose
`python-expert`	Python development, testing, async patterns
`go-expert`	Go development, goroutines, modules
`java-expert`	Java/Spring Boot development
`frontend-expert`	JS/TS/React/Vue/Angular
`bash-expert`	Shell scripting and automation
`docker-expert`	Dockerfile, container orchestration
`kubernetes-expert`	K8s/OpenShift, Helm, GitOps
`jenkins-expert`	CI/CD pipelines, Jenkinsfiles
`git-expert`	Git operations, branching strategies
`test-automator`	Test suites, CI pipelines
`debugger`	Error analysis, debugging
`code-reviewer`	Code quality, security review
`technical-documentation-writer`	Documentation
`api-documenter`	OpenAPI/Swagger specs
`general-purpose`	Fallback for unspecified tasks

Slash Commands

Command	Description
`/github-pr-review`	Review a GitHub PR and post inline comments. Posts as single review with summary.
`/github-review-handler`	Process human reviewer comments from a PR.
`/github-coderabbitai-review-handler`	Process CodeRabbit AI review comments.
`/code-review`	Run code review on local changes.

`/github-pr-review` Features

Auto-detect PR from current branch or accept PR number/URL
Reads project CLAUDE.md for review rules
Deep code review (security, bugs, error handling, performance)
User selection - choose which findings to post
Single review thread with summary table
Inline comments with severity badges, suggestions, AI prompts

MCP Servers

The .claude/servers/ directory contains MCP (Model Context Protocol) server implementations.

Code Execution Server

Located at .claude/servers/code-execution/, this server wraps MCP connections through a code execution layer using UTCP.

Why not connect MCP servers directly?

Approach	API Calls	Flexibility	Performance
Direct MCP	1 call per tool	Single tool per request	Many round-trips
Code Execution	1 call for entire workflow	Chain tools with TypeScript	Single round-trip

Key advantages:

Tool Chaining in One Call

// Direct MCP: 3 separate API calls
// Code Execution: 1 call with this code:
const repos = await github.listRepos();
const metrics = await Promise.all(
  repos.map(r => github.getMetrics(r.id))
);
return summarize(metrics);

Custom Logic Between Tools
- Conditional branching based on results
- Loops and iterations
- Data transformation and filtering
- Error handling and retries
Reduced Token Usage
- One tool call instead of many
- Results aggregated before returning
- No intermediate results in conversation
Better Performance
- Single round-trip to API
- Parallel execution with Promise.all
- No waiting for Claude to process each step

Setup:

cd ~/.claude/servers/code-execution
npm install

Adding Server Configs

Create a config file in .claude/servers/code-execution/configs/:

{
  "manual_call_templates": [
    {
      "name": "my-service",
      "call_template_type": "mcp",
      "config": {
        "mcpServers": {
          "my-server": {
            "transport": "http",
            "url": "http://localhost:8080/mcp"
          }
        }
      }
    }
  ],
  "tool_repository": {
    "tool_repository_type": "in_memory"
  },
  "tool_search_strategy": {
    "tool_search_strategy_type": "tag_and_description_word_match"
  }
}

Add the MCP server to Claude by editing ~/.claude.json:

{
  "mcpServers": {
    "my-service": {
      "command": "npx",
      "args": ["@utcp/code-mode-mcp"],
      "env": {
        "UTCP_CONFIG_FILE": "~/.claude/servers/code-execution/configs/my-service.json"
      }
    }
  }
}

Restart Claude Code for the changes to take effect.

See configs/example.json.example for a config template.

Creating an Agent for Your MCP Server

After adding an MCP server, you'll want a specialized agent to manage it. Ask Claude to create one:

Example prompt:

Create an agent for my new MCP server called "my-service" that handles [describe what your server does].
The agent should be saved to ~/.claude/agents/my-service-manager.md

Agent file structure:

---
name: my-service-manager
description: Use this agent for [your MCP server purpose]
---

You are a specialist agent for managing the my-service MCP server...

## Available Tools
- mcp__my-service__tool_name
- ...

The agent will be automatically available in Claude Code after creation.

Why Agent-Based Workflow?

This configuration implements an orchestrator pattern where Claude acts as a manager delegating to specialist agents. Here's why:

Context Preservation

Main conversation stays lean - The orchestrator only tracks high-level progress
Specialists work in isolation - Each agent handles its task without bloating the main context
Parallel execution - Multiple agents can work simultaneously on independent tasks

Expertise Separation

Domain knowledge - Each agent has specialized instructions for its domain (Python best practices, Git workflows, Docker patterns)
Tool restrictions - Agents only use tools relevant to their specialty
Consistent patterns - Same agent = same coding style and conventions

Quality Assurance

Mandatory code review - Every code change goes through code-reviewer
Automated testing - test-automator runs after changes
Review loop - Changes iterate until approved

Workflow Example

User: "Add a new feature to handle user auth"
         │
         ▼
   ┌─────────────────┐
   │  ORCHESTRATOR   │  (main Claude - manages, doesn't code)
   └────────┬────────┘
            │
   ┌────────┼────────────────┐
   │        │                │
   ▼        ▼                ▼
┌──────┐ ┌──────┐      ┌──────────┐
│python│ │ git  │      │test-auto │
│expert│ │expert│      │  mator   │
└──────┘ └──────┘      └──────────┘
   │        │                │
   └────────┴────────────────┘
            │
            ▼
    ┌──────────────┐
    │code-reviewer │
    └──────────────┘
            │
            ▼
       ✅ Done

Benefits

Reduced token usage - Specialist context is discarded after task completion
Better code quality - Domain experts produce better code
Faster execution - Parallel agents vs sequential operations
Maintainability - Easy to add/modify agents for new domains

How It Works

The CLAUDE.md file defines an orchestrator pattern where:

The main Claude instance acts as a manager/orchestrator
It delegates tasks to specialist agents based on the domain
After code changes, automatic code review and test loops run

Customization

Edit CLAUDE.md to customize orchestrator rules and agent routing
Edit .claude/settings.json to modify hooks and allowed tools
Edit agents in .claude/agents/ to customize agent behavior

File Structure

~/.claude/
├── agents/           # Specialist agent definitions
├── scripts/          # Helper scripts for hooks
├── settings.json     # Hooks and tool permissions
└── statusline.sh     # Status line script

License

MIT

Name		Name	Last commit message	Last commit date
Latest commit History 44 Commits
agents		agents
commands		commands
scripts		scripts
servers/code-execution		servers/code-execution
state		state
.gitignore		.gitignore
CLAUDE.md		CLAUDE.md
README.md		README.md
settings.json		settings.json
statusline.sh		statusline.sh

myk-org/claude-code-config

Folders and files

Latest commit

History

Repository files navigation

Claude Code Config

Requirements

Installation

Option 1: Clone directly to ~/.claude

Option 2: Clone as a regular git repo

Backup Existing Config

Symlink Approach

Copy Approach

Integration with Dotfiles (GNU Stow)

How It Works

Directory Structure

Setup

Updating

What's Included

Agents

Slash Commands

/github-pr-review Features

MCP Servers

Code Execution Server

Adding Server Configs

Creating an Agent for Your MCP Server

Why Agent-Based Workflow?

Context Preservation

Expertise Separation

Quality Assurance

Workflow Example

Benefits

How It Works

Customization

File Structure

License

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Languages

`/github-pr-review` Features

Packages