Skip to main content

​
Quick Start Guide

Get from zero to a fully functional AI agent team in under 5 minutes.

​
Choose Your Integration Method

Claude Code

Recommended for seamless integration

Reference Mode

Use agents as prompt templates
1

Copy agents to your Claude Code directory

# Clone the repository
git clone https://github.com/msitarzewski/agency-agents.git

# Copy agents to your Claude Code directory
cp -r agency-agents/* ~/.claude/agents/
2

Activate agents in your conversations

Simply reference the agent you want to activate in your Claude Code sessions:
Hey Claude, activate Frontend Developer mode and help me build a React component
Or for multiple agents:
Activate Backend Architect and DevOps Automator to help me design and deploy an API
3

Start building

The agent will respond with their specialized personality, workflows, and deliverables.Each agent knows:
  • Their specific role and expertise
  • The workflows they should follow
  • What deliverables they should produce
  • Success metrics for their work

​
Option 2: Use as Reference

Each agent file contains everything you need to use them as prompt templates:
1

Browse the agents

Navigate to the agent directory and find the specialist you need:
agency-agents/
├── engineering/
├── design/
├── marketing/
├── product/
├── project-management/
├── testing/
├── support/
├── spatial-computing/
└── specialized/
2

Read the agent file

Each agent file includes:
  • Identity & Personality - Who they are and how they think
  • Core Mission - What they’re responsible for
  • Critical Rules - Domain-specific guidelines
  • Technical Deliverables - Code examples and templates
  • Workflow Process - Step-by-step approach
  • Success Metrics - How to measure their effectiveness
3

Copy and adapt

Copy the agent content and use it as a prompt in your preferred AI tool:
[Paste agent content]

Now help me [your specific task]

​
Your First Agent

Let’s try activating the Frontend Developer agent: 
# In your Claude Code session:
Hey Claude, activate Frontend Developer mode.

I need to build a data table component in React that:
- Displays 10,000 rows efficiently
- Supports sorting and filtering
- Works on mobile devices
- Follows accessibility best practices

​
Using Multiple Agents

For complex projects, activate multiple agents in sequence or parallel:

​
Sequential Workflow

1

Design Phase

Activate UX Architect to design the component architecture
2

Implementation Phase

Activate Frontend Developer to implement the design
3

Quality Assurance

Activate Evidence Collector to test the implementation
4

Final Review

Activate Reality Checker for production readiness assessment

​
Parallel Workflow

For independent work streams: 
Activate the following agents in parallel:
- Frontend Developer: Build the UI components
- Backend Architect: Design the API
- DevOps Automator: Set up the deployment pipeline

Coordinate their work for a cohesive solution.

​
Real-World Example

Let’s build a startup MVP: 
Activate Agents Orchestrator in NEXUS-Sprint mode.

Project: Task management SaaS MVP
Timeline: 4 weeks

Sprint team:
- PM: Senior Project Manager, Sprint Prioritizer
- Design: UX Architect, Brand Guardian
- Engineering: Frontend Developer, Backend Architect, DevOps Automator
- QA: Evidence Collector, Reality Checker

Begin with architecture and sprint planning.

​
What Agents Say

Each agent has a distinct personality. Here’s what you’ll hear:
“Implemented virtualized table component reducing render time by 80%. Added smooth transitions and micro-interactions for better user engagement. Built with screen reader support and keyboard navigation throughout.”
“Screenshot integration-mobile.png shows broken responsive layout. System needs 2-3 revision cycles before production consideration. Previous claim of ‘luxury design’ not supported by visual evidence.”
“You’re not marketing on Reddit - you’re becoming a valued community member who happens to represent a brand. Success comes from giving more than you take and building genuine relationships over time.”
“Added a celebration animation that reduces task completion anxiety by 40%. Every playful element serves a functional or emotional purpose - design delight that enhances rather than distracts.”

​
Next Steps

Installation Guide

Detailed setup instructions

Agent Design

Understand agent philosophy

Use Cases

Real-world scenarios

Contributing

Create your own agents
Start with a single agent to get familiar with their personality and workflow. Then gradually add more agents as your needs grow.

Build docs developers (and LLMs) love

Get started for freeTalk to us