Claude Code Specialized Agents

A collection of specialized AI agents for Claude Code that enhance software development workflows with focused expertise in specific domains.

Overview

This repository contains intelligent agents designed to work with Claude Code, Anthropic's CLI tool for AI-assisted software development. Each agent is a specialist that brings deep domain knowledge and best practices to help developers build better software faster.

Available Agents

🧠 Cynefin Decision Framework Agent

File: decision-making/cynefin-decision-framework-agent.md

A specialized Task Organization specialist that uses the Cynefin framework to help categorize problems and guide decision-making approaches. This agent excels at:

• Problem Categorization: Sort tasks into appropriate Cynefin domains
• Task Breakdown: Decompose complex tasks into manageable components
• Task Grouping: Organize related tasks by domain and execution strategy
• Execution Strategy: Recommend domain-appropriate approaches for task completion
• Priority Management: Help users sequence tasks based on domain characteristics and dependencies

Perfect for when you're facing complex decisions, unclear problem scope, or need help organizing and prioritizing work based on complexity and context.

🏗️ TDD Chicago School Kotlin Agent

File: testing/tdd-chicago-school-kotlin.md

A Test-Driven Development specialist following the Chicago School (classicist) approach for Kotlin development. This agent emphasizes:

• Inside-Out Development: Start with core domain objects and build outward
• State-Based Verification: Assert on object state rather than interactions
• Emergent Design: Let design emerge naturally from tests
• Minimal Mocking: Use real objects whenever possible
• Kotlin Idioms: Leverage Kotlin's type system and language features

Perfect for developers who want to practice disciplined TDD with a focus on state verification and clean, testable code architecture.

🎨 ASCII UI Prototype Agent

File: design/ui-ascii-prototype.md

A UI/UX design specialist focused on creating ASCII-based wireframes, mockups, and prototypes. This agent excels at:

• ASCII Wireframing: Create clear layout structures using text characters
• Component Visualization: Represent UI elements with consistent ASCII patterns
• Responsive Design Concepts: Show how layouts adapt across different screen sizes
• Interactive Flow Mapping: Visualize user journeys and state changes
• Design Documentation: Create maintainable, text-based design specs

Ideal for rapid prototyping, design documentation that lives alongside code, and communicating UI concepts in text-based environments.

Why Specialized Agents?

Each agent represents years of accumulated best practices and domain expertise, distilled into actionable guidance. Instead of generic AI assistance, you get:

• Domain-Specific Knowledge: Deep understanding of specialized methodologies
• Consistent Practices: Agents follow established patterns and conventions
• Accelerated Learning: Learn best practices while building
• Quality Assurance: Built-in checks and guidance for common pitfalls

How It Works

These agents integrate with Claude Code's agent system. Each .md file defines an agent with:

• Specialized Knowledge: Deep expertise in a specific domain
• Automated Hooks: Pre and post-execution scripts for environment setup
• Consistent Patterns: Established conventions and best practices
• Quality Gates: Built-in validation and verification steps

Getting Started

Prerequisites

• Claude Code - Anthropic's CLI tool for AI-assisted software development
• Basic familiarity with your chosen domain (TDD, UI/UX design, etc.)

Installation & Setup

1. Clone this repository:

   git clone https://github.com/tddworks/claude-agents.git
   cd claude-agents

2. Choose your agent: Copy the agent file you want to use to your project directory or configure Claude Code to reference it.

3. Integration: Each agent file contains metadata that Claude Code uses to understand the agent's capabilities and configure the appropriate environment.

Usage Examples

Using the Cynefin Decision Framework Agent

# When facing complex decisions or unclear problem scope
claude-code --agent decision-making/cynefin-decision-framework-agent.md "Help me organize this complex task list and determine the best approach"

The agent will:

• Categorize problems using the Cynefin framework
• Break down complex tasks into manageable components
• Recommend domain-appropriate execution strategies
• Help prioritize work based on complexity and dependencies
• Guide you through uncertainty with structured decision-making

Using the TDD Chicago School Kotlin Agent

# In your Kotlin project directory
claude-code --agent testing/tdd-chicago-school-kotlin.md "Help me implement a user registration feature using TDD"

The agent will:

• Set up the TDD environment with gradle clean
• Guide you through inside-out development
• Ensure state-based verification patterns
• Run tests and coverage verification
• Follow Kotlin idioms and best practices

Using the ASCII UI Prototype Agent

# In your project directory  
claude-code --agent design/ui-ascii-prototype.md "Create wireframes for a user dashboard"

The agent will:

• Create clear ASCII wireframes
• Show responsive design concepts
• Map interactive flows
• Document design decisions in text format
• Suggest improvements to layout and user experience

Agent Structure

Each agent follows this structure:

---
name: agent-name
type: category
color: "#hexcolor"
description: Brief description
capabilities:
  - capability1
  - capability2
priority: high|medium|low
hooks:
  pre: |
    # Setup commands
  post: |
    # Cleanup/verification commands
---

# Agent content with detailed instructions, examples, and best practices

Contributing

We welcome contributions to expand the collection of specialized agents! Here's how you can help:

Creating a New Agent

1. Identify a Domain: Choose a specific methodology, framework, or development practice that would benefit from specialized guidance.

2. Research Best Practices: Gather established patterns, conventions, and expert knowledge in that domain.

3. Follow the Template: Use the agent structure shown above and include:

   • Clear metadata with appropriate capabilities
   • Comprehensive methodology documentation
   • Practical examples and code samples
   • Common patterns and anti-patterns
   • Integration hooks for environment setup

4. Test Your Agent: Verify it works well with Claude Code and provides valuable, specialized guidance.

Contribution Guidelines

• Focus on Expertise: Agents should represent deep, specialized knowledge rather than general advice
• Include Examples: Provide concrete code examples and usage patterns
• Document Thoroughly: Include both methodology and practical implementation guidance
• Quality over Quantity: Each agent should be comprehensive and well-tested
• Follow Conventions: Use consistent formatting and structure across agents

Submission Process

1. Fork this repository
2. Create a new branch for your agent
3. Add your agent file with thorough documentation
4. Test the agent with Claude Code
5. Submit a pull request with a clear description

Ideas for Future Agents

• DDD (Domain-Driven Design) Specialist: Bounded contexts, aggregates, domain events
• Clean Architecture Guide: Dependency inversion, use cases, adapters
• Security Review Expert: OWASP guidelines, vulnerability assessment
• Performance Optimization: Profiling, bottleneck analysis, optimization strategies
• API Design Specialist: REST, GraphQL, OpenAPI best practices
• Database Design Expert: Normalization, indexing, query optimization

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Anthropic for developing Claude Code
• The open source community for establishing the best practices these agents embody
• Contributors who help expand and improve the agent collection

---

Start building better software with specialized AI guidance. Choose your agent and elevate your development workflow today!