Skip to content

scripts

Directory actions

More options

Directory actions

More options

Latest commit

 

History

History

scripts

README.md

🚀 NANDA Agent Deployment Scripts

Production-ready scripts for deploying NANDA agents to AWS EC2.

📋 Available Scripts

🤖 Single Agent Deployment

aws-single-agent-deployment.sh - Deploy one specialized agent to one EC2 instance

bash aws-single-agent-deployment.sh <AGENT_ID> <API_KEY> <NAME> <DOMAIN> <SPECIALIZATION> <DESCRIPTION> <CAPABILITIES> [REGISTRY_URL] [PORT] [REGION] [INSTANCE_TYPE]

Example:

bash aws-single-agent-deployment.sh \
  "data-scientist" \
  "sk-ant-api03-..." \
  "Data Scientist" \
  "data analysis" \
  "expert data analyst and machine learning specialist" \
  "I help with statistical analysis, machine learning, and data visualization" \
  "python,statistics,machine learning,data visualization" \
  "http://registry.chat39.com:6900" \
  "6000" \
  "us-east-1" \
  "t3.micro"

🏭 Multi-Agent Deployment

aws-multi-agent-deployment.sh - Deploy 10 agents to one EC2 instance

bash aws-multi-agent-deployment.sh <API_KEY> <CONFIG_JSON> [REGISTRY_URL] [REGION] [INSTANCE_TYPE]

Example:

bash aws-multi-agent-deployment.sh \
  "sk-ant-api03-..." \
  "agent_configs/group-01-business-and-finance-experts.json" \
  "http://registry.chat39.com:6900" \
  "us-east-1" \
  "t3.xlarge"

🖥️ Existing Server Deployment

deploy-agent.sh - Deploy agent to existing Ubuntu/Amazon Linux server

bash deploy-agent.sh <AGENT_TYPE> <AGENT_ID> <API_KEY> [PORT] [REGISTRY_URL]

📦 Pre-configured Agent Groups

Ready-to-deploy agent configurations:

File Agents Description
agent_configs/group-01-business-and-finance-experts.json 10 Financial analysts, advisors, strategists
agent_configs/group-02-technology-and-engineering.json 10 Software engineers, DevOps, AI researchers
agent_configs/group-03-creative-and-design.json 10 Designers, content creators, brand experts
agent_configs/group-04-healthcare-and-life-sciences.json 10 Medical researchers, health informatics
agent_configs/group-05-education-and-research.json 10 Academic researchers, educators
agent_configs/group-06-media-and-entertainment.json 10 Journalists, producers, social media
agent_configs/group-07-environmental-and-sustainability.json 10 Climate scientists, sustainability experts
agent_configs/group-08-social-services-and-community.json 10 Social workers, policy analysts
agent_configs/group-09-sports-and-recreation.json 10 Fitness trainers, sports analysts
agent_configs/group-10-travel-and-hospitality.json 10 Travel planners, hospitality managers
agent_configs/100-agents-config.json 100 All agent personalities combined

🛠️ Prerequisites

  • AWS CLI configured with credentials (aws configure)
  • Anthropic API Key for Claude LLM
  • SSH Key Pair for EC2 access (automatically created)

🎯 Recommended Instance Types

Deployment Instance Type Cost Use Case
Single Agent t3.micro $8/month Development, testing
Multi-Agent (10) t3.xlarge $150/month Production, high traffic
High Performance t3.2xlarge $300/month Enterprise, 20+ agents

🔧 What the Scripts Do

  1. 🔐 AWS Setup: Create security groups, key pairs, open ports
  2. 🖥️ EC2 Launch: Launch Ubuntu 22.04 instance with user-data script
  3. 📦 Dependencies: Install Python, git, anthropic library
  4. 📂 Project Setup: Clone repo, create virtual environment
  5. 🤖 Agent Start: Configure and start agent(s) with supervisor
  6. 📋 Registry: Register agent(s) with NANDA registry
  7. ✅ Health Check: Verify agent(s) are responding

🧪 Testing Deployed Agents

Test Single Agent

curl -X POST http://AGENT_IP:6000/a2a \
  -H "Content-Type: application/json" \
  -d '{"content":{"text":"Hello! What are your capabilities?","type":"text"},"role":"user","conversation_id":"test123"}'

Test A2A Communication

curl -X POST http://AGENT_A_IP:6000/a2a \
  -H "Content-Type: application/json" \
  -d '{"content":{"text":"@agent-b-id Can you help with this task?","type":"text"},"role":"user","conversation_id":"test123"}'

🛑 Cleanup

To terminate instances:

# Single agent
aws ec2 terminate-instances --region us-east-1 --instance-ids i-xxxxx

# Multiple instances
aws ec2 describe-instances --filters "Name=tag:Project,Values=NANDA*" --query 'Reservations[*].Instances[*].InstanceId' --output text | xargs aws ec2 terminate-instances --region us-east-1 --instance-ids

🚨 Troubleshooting

Common Issues

Agent not responding:

  • Check security group has port open
  • Verify agent process is running: ps aux | grep python
  • Check logs: tail -f agent.log

SSH connection failed:

  • Ensure using correct .pem key file
  • Check key permissions: chmod 400 *.pem
  • Verify instance is running: aws ec2 describe-instances

Registration failed:

  • Verify registry URL is accessible
  • Check public IP retrieval in user-data logs
  • Ensure ANTHROPIC_API_KEY is valid

Debug Commands

# SSH into instance
ssh -i nanda-agent-key.pem ubuntu@INSTANCE_IP

# Check user-data logs
sudo tail -f /var/log/cloud-init-output.log

# Check agent logs
cd nanda-agent-* && tail -f agent.log

# Check running processes
ps aux | grep python

📈 Scaling

Deploy 100 Agents (10 instances)

for i in {1..10}; do
  bash aws-multi-agent-deployment.sh \
    "sk-ant-api03-..." \
    "group-0${i}-*.json" \
    "http://registry.chat39.com:6900" \
    "us-east-1" \
    "t3.xlarge" &
done

Cross-Region Deployment

# Deploy to multiple regions
for region in us-east-1 us-west-2 eu-west-1; do
  bash aws-multi-agent-deployment.sh \
    "sk-ant-api03-..." \
    "group-01-business-and-finance-experts.json" \
    "http://registry.chat39.com:6900" \
    "$region" \
    "t3.xlarge" &
done

🎯 Ready to deploy? Start with a single agent to test, then scale to multi-agent deployments!