agents/README.md

Claude Code Workflows & Agents

A comprehensive production-ready system combining 83 specialized AI agents, 15 multi-agent workflow orchestrators, and 42 development tools for Claude Code.

⚠️ Major Update: This repository has been restructured. If you're upgrading from a previous version, see the Migration Guide below.

Quick Links

• Installation - New users start here
• Migration Guide - Upgrading from previous version? Read this first
• Workflow Commands - Multi-agent orchestration
• Development Tools - Single-purpose utilities
• Agent Categories - All 83 agents organized by domain

Overview

This unified repository provides everything needed for intelligent automation and multi-agent orchestration across modern software development:

• 83 Specialized Agents - Domain experts with deep knowledge (architecture, languages, infrastructure, quality, data/AI, business)
• 15 Workflow Orchestrators - Multi-agent coordination systems for complex operations
• 42 Development Tools - Focused utilities for specific tasks

System Requirements

• Claude Code installed and configured
• Git for repository management

Installation

New Installation

cd ~/.claude
git clone https://github.com/wshobson/agents.git

All agents, workflows, and tools will be automatically available to Claude Code.

Updating from Previous Version

If you previously had the agents repository installed:

cd ~/.claude/agents
git pull origin main

Important: The repository structure has changed. All agent files have been moved to the agents/ subdirectory. Claude Code will automatically detect the new structure.

Repository Structure

agents/
├── agents/                        # 83 specialized AI agents
│   ├── backend-architect.md
│   ├── frontend-developer.md
│   └── ... (all agent definitions)
├── workflows/                     # 15 multi-agent orchestrators
│   ├── feature-development.md
│   ├── full-stack-feature.md
│   ├── security-hardening.md
│   └── ... (workflow commands)
├── tools/                         # 42 development utilities
│   ├── api-scaffold.md
│   ├── security-scan.md
│   └── ... (tool commands)
└── README.md

Workflow Commands

Multi-agent orchestration systems that coordinate complex, cross-domain tasks:

Core Development Workflows

Command	Purpose	Agent Coordination
feature-development	End-to-end feature implementation	Backend, frontend, testing, deployment
full-stack-feature	Complete multi-tier implementation	Backend API, frontend UI, mobile, database
full-review	Multi-perspective code analysis	Architecture, security, performance, quality
smart-fix	Intelligent problem resolution	Dynamic agent selection based on issue type
tdd-cycle	Test-driven development orchestration	Test writer, implementer, refactoring specialist

Process Automation Workflows

Command	Purpose	Scope
git-workflow	Version control process automation	Branching strategies, commit standards, PR templates
improve-agent	Agent optimization	Prompt engineering, performance tuning
legacy-modernize	Codebase modernization	Architecture migration, dependency updates
multi-platform	Cross-platform development	Web, mobile, desktop coordination
workflow-automate	CI/CD pipeline automation	Build, test, deploy, monitor

Advanced Orchestration Workflows

Command	Primary Focus	Specialized Agents
security-hardening	Security-first development	Threat modeling, vulnerability assessment
data-driven-feature	ML-powered functionality	Data science, feature engineering, model deployment
ml-pipeline	End-to-end ML infrastructure	MLOps, data engineering, model serving
performance-optimization	System-wide optimization	Profiling, caching, query optimization
incident-response	Production issue resolution	Diagnostics, root cause analysis, hotfix deployment

Development Tools

Focused, single-purpose utilities for specific development operations:

AI and Machine Learning

• langchain-agent - LangChain agent development
• ai-assistant - AI-powered development assistance
• ai-review - AI-based code review

API Development

• api-scaffold - API endpoint scaffolding
• api-mock - API mocking and testing

Testing & Quality

• tdd-red - Red phase (failing tests)
• tdd-green - Green phase (passing implementation)
• tdd-refactor - Refactor phase
• test-harness - Test infrastructure setup

Security & Compliance

• security-scan - Vulnerability scanning
• compliance-check - Compliance validation

Infrastructure & Operations

• k8s-manifest - Kubernetes manifest generation
• docker-optimize - Docker optimization
• monitor-setup - Monitoring infrastructure
• deploy-checklist - Deployment validation

Code Quality

• code-explain - Code explanation
• code-migrate - Code migration
• refactor-clean - Code refactoring
• pr-enhance - Pull request enhancement

And 20+ more tools for debugging, documentation, data validation, cost optimization, and developer workflows

Usage

All 83 specialized agents are automatically available to Claude Code and can be explicitly invoked:

"Use backend-architect to design the authentication API"
"Have security-auditor scan for OWASP vulnerabilities"
"Get performance-engineer to optimize this bottleneck"

Agent Categories

Architecture & System Design (7 agents)

backend-architect, cloud-architect, kubernetes-architect, hybrid-cloud-architect, graphql-architect, terraform-specialist, architect-review

Programming Languages (15 agents)

javascript-pro, typescript-pro, python-pro, golang-pro, rust-pro, java-pro, csharp-pro, c-pro, cpp-pro, ruby-pro, php-pro, scala-pro, elixir-pro, django-pro, fastapi-pro

Infrastructure & Operations (9 agents)

devops-troubleshooter, deployment-engineer, database-admin, database-optimizer, database-architect, network-engineer, incident-responder, performance-engineer, observability-engineer

Security & Quality (9 agents)

code-reviewer, security-auditor, backend-security-coder, frontend-security-coder, mobile-security-coder, test-automator, tdd-orchestrator, debugger, error-detective

Frontend & Mobile (7 agents)

frontend-developer, ui-ux-designer, ui-visual-validator, mobile-developer, ios-developer, flutter-expert, unity-developer

Data & AI (6 agents)

data-scientist, data-engineer, ml-engineer, mlops-engineer, ai-engineer, prompt-engineer

Documentation (5 agents)

docs-architect, api-documenter, reference-builder, tutorial-engineer, mermaid-expert

Business & Operations (6 agents)

business-analyst, hr-pro, legal-advisor, customer-support, sales-automator, content-marketer

SEO & Content (10 agents)

seo-content-writer, seo-content-auditor, seo-keyword-strategist, seo-meta-optimizer, seo-structure-architect, seo-snippet-hunter, seo-content-refresher, seo-cannibalization-detector, seo-authority-builder, seo-content-planner

Specialized Domains (7 agents)

blockchain-developer, quant-analyst, risk-manager, payment-integration, minecraft-bukkit-pro, legacy-modernizer, context-manager

Utilities (3 agents)

search-specialist, dx-optimizer, sql-pro

Multi-Agent Orchestration Examples

Agents automatically coordinate for complex tasks:

Full-Stack Development

Example: "Implement user dashboard with analytics" Orchestrates: backend-architect → graphql-architect → frontend-developer → mobile-developer → test-automator → security-auditor → performance-engineer → deployment-engineer

Security Hardening

Example: "Implement security best practices" Orchestrates: security-auditor → backend-security-coder → frontend-security-coder → mobile-security-coder → test-automator

Data/ML Pipeline

Example: "Build customer churn prediction model" Orchestrates: data-scientist → data-engineer → ml-engineer → mlops-engineer → ai-engineer → performance-engineer

Incident Response

Example: "Debug high CPU usage in production" Orchestrates: incident-responder → devops-troubleshooter → debugger → error-detective → observability-engineer

Model Configuration

Agents are assigned to specific Claude models based on task complexity:

Model	Count	Use Cases
Opus	22	Complex architecture, critical analysis, security audits, business operations
Sonnet	50	Standard development, engineering tasks, quality assurance
Haiku	11	Quick focused tasks, SEO optimization, reference building

Multi-Agent Orchestration Patterns

Sequential Processing

backend-architect → frontend-developer → test-automator → security-auditor

Parallel Execution

performance-engineer + database-optimizer → Merged optimization

Conditional Routing

debugger → [backend-architect | frontend-developer | devops-troubleshooter]

Validation Pipeline

feature-development → security-auditor → performance-engineer → Validated release

Migration Guide

What Changed?

Major Update: This repository has been restructured to consolidate all Claude Code extensions in one place:

1. Repository Structure: All agents moved from root to agents/ subdirectory
2. Workflows Added: 15 multi-agent workflow orchestrators (previously in separate commands repo)
3. Tools Added: 42 development utilities (previously in separate commands repo)
4. Unified Experience: Everything now accessible from a single repository

Migrating from Commands Repository

If you previously used the separate commands repository (wshobson/commands):

Before (old structure):

~/.claude/
├── agents/              # Agents repository
│   └── *.md files
└── commands/            # Commands repository (DEPRECATED)
    └── *.md files

After (new unified structure):

~/.claude/
└── agents/              # Unified repository
    ├── agents/
    ├── workflows/
    └── tools/

Migration Steps

1. Update the agents repository:

   cd ~/.claude/agents
   git pull origin main
   

2. Remove the old commands repository (if installed):

   rm -rf ~/.claude/commands
   

3. Verify installation:

   ls ~/.claude/agents
   # Should show: agents/ workflows/ tools/ README.md
   

What Stays the Same?

• All 83 agents work exactly as before (no command syntax changes)
• Agent definitions and capabilities unchanged
• Direct agent invocation still works: "Use backend-architect to..."

What's New?

✅ Workflows & Tools Added: 15 workflow orchestrators and 42 development tools now included ✅ Unified Repository: Everything accessible from a single installation ✅ Comprehensive Documentation: Migration guide and detailed usage instructions

Need Help?

For detailed migration instructions and troubleshooting, see MIGRATION.md.

If you encounter issues after migrating:

1. Verify directory structure: ls -la ~/.claude/agents
2. Check git status: cd ~/.claude/agents && git status
3. Review common issues in MIGRATION.md
4. Report issues at: https://github.com/wshobson/agents/issues

Contributing

To add new agents, workflows, or tools:

1. Place agent definitions in agents/ directory
2. Place workflow orchestrators in workflows/ directory
3. Place tool commands in tools/ directory
4. Follow existing naming conventions (lowercase, hyphen-separated)
5. Include proper frontmatter in markdown files

License

MIT License - see LICENSE file for details.

Resources

• Claude Code Documentation
• Subagents Documentation
• Claude Code GitHub