Complete Workshop Guide

Your comprehensive guide to the Multi-Agent System Workshop—what’s included, how to use it, and how to deploy it.

🎉 What You Have

A complete 90-minute hands-on workshop for building production-ready multi-agent systems using the Strands framework, presented through a beautiful, modern website.

Workshop Components

📚 Learning Materials:

🌐 Interactive Website:

💻 Working Code:

📊 Workshop Structure

Timeline (90 minutes)

TimeStepTopicGit Tag
0-5minIntroWelcome & Setup-
5-15minStep 1Script Agentv0.1.0
15-30minStep 2Social Agent (Built-in Tools)v0.3.0
30-40minStep 3TTS Agent (Custom Tools)v0.4.0
40-55minStep 4Research Agentv0.5.0
55-70minStep 5Orchestratorv0.7.0
70-80minStep 6Testing & Productionv0.7.0
80-90minWrap-upQ&A & Next Steps-

Learning Progression

Step 1: Agent Basics
  ↓ Learn: initialization, models, prompts
  
Step 2: Built-in Tools  
  ↓ Learn: tool integration, fetch_url
  
Step 3: Custom Tools
  ↓ Learn: external APIs, ElevenLabs
  
Step 4: Complex Workflows
  ↓ Learn: retries, token mgmt, Tavily
  
Step 5: Orchestration
  ↓ Learn: multi-agent coordination
  
Step 6: Production
  ↓ Learn: testing, deployment
  
Complete Multi-Agent System! 🎉

What Participants Build

Complete Pipeline Architecture

Output: Podcast Script + Social Post + Audio File

🚀 Getting Started

For Participants

1. Clone the Repository

git clone https://github.com/radixia/rooting-agentic-pipeline.git
cd rooting-agentic-pipeline

2. Install Dependencies

# Using uv (recommended)
uv sync

# Or using pip
pip install -e .

3. Set Up API Keys

Create a .env file:

cp .env.template .env

Add your API keys:

ANTHROPIC_API_KEY=your_key_here
OPENAI_API_KEY=your_key_here
TAVILY_API_KEY=your_key_here
ELEVENLABS_API_KEY=your_key_here

4. View the Workshop Website

cd workshop/site
npm install
npm run dev

Visit http://localhost:4321

5. Start Learning

Navigate through the steps in order, checking out the corresponding git tags:

git checkout v0.1.0  # Step 1
git checkout v0.3.0  # Step 2
# etc...

For Instructors

1. Review All Materials

2. Prepare Environment

3. Pre-Workshop Checklist

4. During Workshop

🤖 Model Configuration & Costs

Default Models

The workshop uses carefully selected models to balance performance and cost:

Anthropic Claude Sonnet 4.5 (claude-sonnet-4-5-20250929)

OpenAI GPT-5 Mini (gpt-5-mini-2025-08-07)

💡 Cost Savings Tip: GPT-5 Mini provides substantial cost savings (60-80% cheaper than GPT-4o) without sacrificing quality for tasks like social media posts. This makes the workshop more affordable to run while still producing professional results.

Customizing Models

You can override the default models via environment variables:

# Use different models
export ANTHROPIC_MODEL_ID="claude-opus-4-20250514"
export OPENAI_MODEL_ID="gpt-4o"

Or in your .env file:

ANTHROPIC_MODEL_ID=claude-opus-4-20250514
OPENAI_MODEL_ID=gpt-4o

Cost Comparison

Approximate costs per 1M tokens (input/output):

ModelInputOutputBest For
Claude Sonnet 4.5$3$15Complex reasoning, long content
Claude Opus 4$15$75Maximum quality
GPT-5 Mini$0.15$0.60Cost-efficient tasks
GPT-4o$2.50$10Balanced performance

Workshop Cost Estimate: Running the complete pipeline (research + script + social + audio) typically costs $0.15-0.30 per execution with default models.

🌐 Website Features

Design

Dark Theme: Professional color scheme optimized for developers

Typography: System font stack for optimal readability

Code Highlighting: GitHub Dark theme with proper syntax support

Responsive: Fixed sidebar on desktop, mobile-friendly layout

The sidebar provides quick access to:

Workshop Steps:

Resources:

Technical Stack

🛠️ Development

Website Commands

cd workshop/site

# Install dependencies
npm install

# Start dev server (http://localhost:4321)
npm run dev

# Build for production
npm run build

# Preview production build
npm run preview

Customizing the Website

Change Colors

Edit CSS variables in src/layouts/Layout.astro:

:root {
  --color-bg: #0f172a;           /* Background */
  --color-surface: #1e293b;      /* Cards/surfaces */
  --color-accent: #3b82f6;       /* Links/highlights */
  --color-text: #e2e8f0;         /* Main text */
  --color-text-dim: #94a3b8;     /* Secondary text */
}

Add Navigation Items

Update sidebar in src/layouts/Layout.astro:

<ul class="nav-links">
  <li><a href="/new-page" class="nav-link">New Page</a></li>
</ul>

Create New Pages

Add MDX files to src/pages/:

---
layout: ../layouts/Layout.astro
title: "My New Page"
---

# My New Page

Content here...

Modify Styles

Edit the <style> section in src/layouts/Layout.astro for global styles.

🚢 Deployment

Quick Deploy:

npm install -g vercel
cd workshop/site
vercel

Automatic Deployments:

  1. Push to GitHub
  2. Connect repository to Vercel
  3. Auto-deploy on every push

Netlify

Via CLI:

cd workshop/site
npm run build
netlify deploy --prod --dir=dist

Via UI:

  1. Build command: npm run build
  2. Publish directory: dist
  3. Deploy!

GitHub Pages

Setup:

# .github/workflows/deploy.yml
name: Deploy
on:
  push:
    branches: [main]
jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - uses: actions/setup-node@v3
      - run: cd workshop/site && npm install && npm run build
      - uses: peaceiris/actions-gh-pages@v3
        with:
          github_token: ${{ secrets.GITHUB_TOKEN }}
          publish_dir: ./workshop/site/dist

AWS S3 + CloudFront

cd workshop/site
npm run build

# Upload to S3
aws s3 sync dist/ s3://your-bucket-name/

# Invalidate CloudFront cache
aws cloudfront create-invalidation \
  --distribution-id YOUR_DIST_ID \
  --paths "/*"

Docker

# Dockerfile
FROM node:20-alpine AS builder
WORKDIR /app
COPY workshop/site/package*.json ./
RUN npm install
COPY workshop/site .
RUN npm run build

FROM nginx:alpine
COPY --from=builder /app/dist /usr/share/nginx/html
EXPOSE 80
CMD ["nginx", "-g", "daemon off;"]

Build and run:

docker build -t workshop-site .
docker run -p 8080:80 workshop-site

📈 Workshop Statistics

Content Volume

Learning Outcomes

By completing this workshop, participants will:

✅ Understand agent architecture patterns
✅ Build production-ready AI agents
✅ Integrate tools (built-in and custom)
✅ Implement error handling and retries
✅ Manage tokens and conversations
✅ Orchestrate multi-agent workflows
✅ Deploy and monitor systems
✅ Apply best practices

Time Investment

Self-paced learning: ~3-4 hours

Guided workshop: 90 minutes

Post-workshop exploration: Unlimited!

🎯 Use Cases

For Companies

For Educators

For Individuals

🔧 Troubleshooting

Website Issues

Port already in use?

lsof -ti:4321 | xargs kill -9

Build errors?

rm -rf node_modules package-lock.json .astro
npm install

Styles not updating?

rm -rf .astro
npm run dev

Workshop Issues

Agent initialization fails?

Import errors?

uv sync
# or
pip install -e .

Rate limits?

Token limits?

See the Troubleshooting Guide for 50+ more issues and solutions.

📚 Additional Resources

Documentation

Workshop Materials

External Resources

🤝 Contributing

Improve the Workshop

Found a typo? Have a suggestion? Want to add content?

For Content Issues:

  1. Edit MDX files in workshop/site/src/pages/
  2. Test locally with npm run dev
  3. Submit a pull request

For Website Issues:

  1. Edit layouts or styles in src/
  2. Test thoroughly
  3. Submit PR with screenshots

For Code Issues:

  1. Fix in main repository
  2. Update workshop steps if needed
  3. Test all affected steps

Share Your Experience

After running the workshop:

📄 License

This workshop is part of the rooting-pipeline project. See repository license for details.

🎓 Acknowledgments

Built with:

Inspired by:

🚀 Quick Links

Ready to build multi-agent systems? Begin Step 1 →