Claude Config

Personal Claude Code configuration with custom agents, commands, skills, hooks, and MCP server setup.

Overview

This repository contains my production Claude Code configuration featuring:

• 27 specialized agents for different domains (backend, frontend, DevOps, QA, etc.)
• 26+ custom commands for workflows like analysis, design, testing, research
• Custom skills for Linear, GitHub workflows, monitoring, and more
• MCP server integrations (Serena, Context7, Playwright, Linear, Memory, etc.)
• Event hooks for automation

Quick Start

# Clone to ~/.claude
git clone https://github.com/Kblack0610/claude-config.git ~/.claude

# Or as a dotfiles submodule
cd ~/.dotfiles
git submodule add https://github.com/Kblack0610/claude-config.git .claude

# Restart Claude Code
claude

Structure

claude-config/
├── agents/              # Specialized agent definitions (27 agents)
├── commands/            # Custom slash commands (26+ commands)
│   ├── sc/             # Skill commands (/sc:*)
│   ├── kb/             # Knowledge base commands (/kb:*)
│   ├── daily/          # Daily workflow commands
│   └── my/             # Personal utility commands
├── skills/              # Reusable skill modules
├── hooks/               # Event hooks (prompt submit, etc.)
├── mcp.json            # MCP server configuration
├── settings.json       # Claude Code settings
└── setup-personal-mcp.sh  # MCP server installer

Agents

Specialized Domain Agents

Agent	Purpose
backend-architect	Design reliable backend systems with data integrity focus
frontend-architect	Create accessible, performant UIs with modern frameworks
devops-architect	Automate infrastructure and deployment processes
system-architect	Design scalable system architecture
security-engineer	Identify vulnerabilities and ensure compliance
performance-engineer	Optimize system performance through measurement
quality-engineer	Comprehensive testing strategies
refactoring-expert	Improve code quality and reduce technical debt
python-expert	Production-ready Python code following best practices

Analysis & Planning Agents

Agent	Purpose
deep-research-agent	Comprehensive research with adaptive strategies
requirements-analyst	Transform ideas into concrete specifications
root-cause-analyst	Systematically investigate complex problems
business-panel-experts	Multi-expert business strategy panel

Learning & Mentorship Agents

Agent	Purpose
learning-guide	Teach programming with progressive learning
socratic-mentor	Educational guide using Socratic method
technical-writer	Create clear, comprehensive documentation

Knowledge Base (KB) Agents

Agent	Purpose
kb-architect	Transform product briefs into technical specs
kb-developer	Implement specs with production-ready code
kb-product-owner	Define product direction and create briefs
kb-qa	Verify PRs meet quality gates
kb-linear-implementer	Implement Linear tickets with isolated context

Commands

Analysis Commands (/sc:analyze)

• /sc:analyze - Comprehensive code analysis across quality, security, performance
• /sc:brainstorm - Interactive requirements discovery through Socratic dialogue
• /sc:explain - Explain code, concepts, and system behavior

Design & Architecture (/sc:design)

• /sc:design - Design system architecture, APIs, component interfaces
• /sc:workflow - Generate structured implementation workflows from PRDs
• /sc:spec-panel - Multi-expert specification review

Implementation (/sc:implement, /sc:build)

• /sc:implement - Feature implementation with intelligent persona activation
• /sc:build - Build, compile, and package projects with error handling
• /sc:cleanup - Systematically clean up code and optimize structure
• /sc:improve - Apply systematic improvements to code quality

Testing & Quality (/sc:test, /sc:troubleshoot)

• /sc:test - Execute tests with coverage analysis
• /sc:troubleshoot - Diagnose and resolve issues in code/builds/deployments
• /sc:manual-test - Manual PR testing using Playwright MCP

Documentation (/sc:document, /sc:index)

• /sc:document - Generate focused documentation for components/functions/APIs
• /sc:index - Generate comprehensive project documentation
• /sc:help - List all available /sc commands

Research & Estimation

• /sc:research - Deep web research with adaptive planning
• /sc:estimate - Provide development estimates for tasks/features

Knowledge Base Workflows (/kb:*)

• /kb:ticket - Implement a Linear ticket with CI checks and PR creation
• /kb:implement - Implement a feature and create a PR
• /kb:workflow - Run full G2I workflow (brief → spec → code → review)
• /kb:ci - Quick CI checks (linting, typecheck, format)
• /kb:ci-analyze - Comprehensive CI analysis with E2E tests

Daily Workflows (/daily:*)

• /daily:standup - Generate concise standup update (yesterday, today, blockers)
• /daily:summary - Comprehensive daily activity summary (Linear, GitHub, commits)
• /daily:analysis - Comprehensive repo analysis with auto-fixes and Linear tickets
• /daily:slack - Generate and optionally post daily summary to Slack
• /daily:help - List all /daily commands

Personal Utilities (/my:*)

• /my:kill-all - Kill all processes running for manual testing
• /my:create - Create and run a new skill from a prompt on the fly
• /my:pr-merge-flow - Review PR, fix issues, monitor CI, merge
• /my:monitor-pr - Monitor PR status and CI checks until completion
• /my:claude-edit - Edit Claude configuration files in ~/.dotfiles/.claude

MCP Servers

Pre-configured integrations:

Server	Purpose
serena	Semantic code understanding and symbol manipulation
context7	Up-to-date library documentation and examples
playwright	Browser automation and testing
linear	Linear issue management and workflows
memory	Knowledge graph for project context
sequential-thinking	Chain-of-thought problem solving
github-gh	GitHub CLI integration
filesystem	Enhanced file operations
docker-mcp	Docker container management

MCP Setup

# Auto-install configured MCP servers
./setup-personal-mcp.sh

# Or manually review mcp.json and install servers
cat mcp.json

Skills

Reusable skill modules for complex workflows:

• Linear ticket implementation
• GitHub PR workflows
• Daily repository analysis
• CI/CD automation
• Manual testing automation

Hooks

Event-triggered automation:

• Prompt submit hooks for pre-processing
• Response hooks for post-processing
• Error hooks for debugging

Configuration

settings.json

Claude Code settings including:

• Default model selection
• Token limits
• Auto-save preferences
• UI customizations

Customization

Adding Your Own Agents

---
name: my-custom-agent
description: What this agent does
model: sonnet
---

# System Prompt

Your agent's system prompt here...

Save to agents/my-custom-agent.md

Adding Custom Commands

Create commands/my-command.md:

---
name: my-command
description: What this command does
---

Prompt for the command...

Use with /my-command

Adding Skills

Create skills/my-skill.md with frontmatter:

---
name: my-skill
description: Skill description
compat:
  binks_agent: true
io:
  input: "Input description"
  output: "Output description"
tools:
  - tool1
  - tool2
---

Skill implementation...

Tips

• Agent Selection: Use specialized agents for domain-specific work
• Command Chaining: Combine commands for complex workflows
• MCP Integration: Leverage MCP servers for external tool access
• Skill Reuse: Build skills for frequently repeated tasks

Advanced Usage

Multi-Agent Workflows

# Use kb-product-owner to write brief
# Then kb-architect to create spec
# Then kb-developer to implement
# Finally kb-qa to review

Custom Hooks

Add hooks to hooks/ for automated workflows on events.

Contributing

This is my personal config, but feel free to:

• Fork and customize
• Suggest improvements via issues
• Share your own agent/command/skill designs

License

MIT - Use freely, adapt to your needs.

Credits

Built for Claude Code power users who want maximum productivity.

Inspired by the Claude Code community and various productivity workflows.