Deploy: GCP

Deploy agents to Google Cloud Run using praisonai deploy

Deploy agents to Google Cloud Run using praisonai deploy run --type cloud --provider gcp.

Quick Start - CLI

1. Install PraisonAI

   pip install praisonai
   
2. Set Environment Variables

   export OPENAI_API_KEY="your-key"
   export GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account.json"
   
3. Initialize GCP Config

   praisonai deploy init --file agents.yaml --type cloud --provider gcp
   
4. Check Readiness

   praisonai deploy doctor --provider gcp
   
5. Deploy to GCP

   praisonai deploy run --file agents.yaml --type cloud --provider gcp
   

   Expected Output:

   🚀 Starting deployment...
   
   ☁️  Deploying to Google Cloud Run
     • Project: your-project-id
     • Region: us-central1
     • Service: praisonai-service
   
   📦 Building and pushing Docker image...
   🚀 Deploying to Cloud Run...
   
   ✅ Deployment successful!
   🔗 URL: https://praisonai-service-abc123-uc.a.run.app
   
   Metadata:
     • provider: gcp
     • project_id: your-project-id
     • region: us-central1
     • service: praisonai-service
   
6. Verify

   curl https://praisonai-service-abc123-uc.a.run.app/health
   

Quick Start - SDK

1. Create Deploy Script

   from praisonai.deploy import Deploy, DeployConfig, DeployType, CloudProvider
   from praisonai.deploy.models import CloudConfig
   
   config = DeployConfig(
       type=DeployType.CLOUD,
       cloud=CloudConfig(
           provider=CloudProvider.GCP,
           region="us-central1",
           service_name="praisonai-service",
           project_id="your-project-id",
           cpu="1",
           memory="512",
           min_instances=0,
           max_instances=10
       )
   )
   
   deploy = Deploy(config, "agents.yaml")
   result = deploy.deploy()
   
   print(f"URL: {result.url}")
   
2. Run

   python deploy_gcp.py
   

agents.yaml with GCP Config

framework: praisonai
topic: helpful assistant
roles:
  assistant:
    role: Assistant
    goal: Help users with their questions
    backstory: You are a helpful assistant
    tasks:
      help_task:
        description: Answer user questions
        expected_output: Helpful response

deploy:
  type: cloud
  cloud:
    provider: gcp
    region: "us-central1"
    service_name: "praisonai-service"
    project_id: "${GOOGLE_CLOUD_PROJECT}"
    cpu: "1"
    memory: "512"
    min_instances: 0
    max_instances: 10
    env_vars:
      OPENAI_API_KEY: "${OPENAI_API_KEY}"

GCP Cloud Config Options

Field	Type	Default	Description
provider	string	-	Must be gcp
region	string	-	GCP region (e.g., us-central1)
service_name	string	-	Cloud Run service name
project_id	string	-	GCP project ID
cpu	string	1	CPU allocation
memory	string	512	Memory in MB
min_instances	int	0	Minimum instances (0 = scale to zero)
max_instances	int	10	Maximum instances
env_vars	dict	null	Environment variables

Check Deployment Status

praisonai deploy status --file agents.yaml

Destroy Deployment

praisonai deploy destroy --file agents.yaml --yes

Troubleshooting

Issue	Fix
GCP credentials	gcloud auth login or set GOOGLE_APPLICATION_CREDENTIALS
Project not set	Add project_id to agents.yaml
Permission denied	Check IAM permissions for Cloud Run
Deploy failed	Run praisonai deploy doctor --provider gcp

Manual GCP CLI Deployment (Optional)

These commands are for manual deployment only. Use praisonai deploy for automated deployment.

Never pass secrets via --set-env-vars — they appear in ps, shell history, and CI logs. Use --env-vars-file with chmod 0600. The praisonai deploy run command does this automatically (since PR #1744).

# Initialize and enable services
gcloud init
gcloud services enable run.googleapis.com
gcloud services enable containerregistry.googleapis.com
gcloud services enable cloudbuild.googleapis.com

# Set environment variables
export OPENAI_MODEL_NAME="gpt-4o"
export OPENAI_API_KEY="your-api-key"
export OPENAI_API_BASE="https://api.openai.com/v1"

# Configure Docker authentication
yes | gcloud auth configure-docker us-central1-docker.pkg.dev 
gcloud artifacts repositories create praisonai-repository --repository-format=docker --location=us-central1

# Build and push Docker image
PROJECT_ID=$(gcloud config get-value project)
TAG="latest"
docker build --platform linux/amd64 -t gcr.io/${PROJECT_ID}/praisonai-app:${TAG} .
docker tag gcr.io/${PROJECT_ID}/praisonai-app:${TAG} us-central1-docker.pkg.dev/${PROJECT_ID}/praisonai-repository/praisonai-app:${TAG}
docker push us-central1-docker.pkg.dev/${PROJECT_ID}/praisonai-repository/praisonai-app:${TAG}

# Create secure env vars file (NEVER use --set-env-vars for secrets)
cat > /tmp/praisonai-env.yaml <<EOF
OPENAI_MODEL_NAME: "${OPENAI_MODEL_NAME}"
OPENAI_API_KEY: "${OPENAI_API_KEY}"
OPENAI_API_BASE: "${OPENAI_API_BASE}"
EOF
chmod 600 /tmp/praisonai-env.yaml

# Deploy to Cloud Run with secure env vars file
gcloud run deploy praisonai-service \
    --image us-central1-docker.pkg.dev/${PROJECT_ID}/praisonai-repository/praisonai-app:${TAG} \
    --platform managed \
    --region us-central1 \
    --allow-unauthenticated \
    --env-vars-file /tmp/praisonai-env.yaml

# Clean up temp file
rm /tmp/praisonai-env.yaml

Security: How praisonai deploy Handles Secrets

The praisonai deploy run command automatically handles secrets securely:

• Creates temp YAML via tempfile.mkstemp() with mode 0600 (owner read/write only)
• Passes secrets via --env-vars-file instead of command line arguments
• Removes temp file in finally block, even if deployment fails
• No secrets appear in ps, shell history, or CI logs

Validation Commands:

# Verify GCP CLI
gcloud config get-value project

# List Cloud Run services
gcloud run services list

# Describe service
gcloud run services describe praisonai-service --region us-central1

# View logs
gcloud run services logs read praisonai-service --region us-central1

Related

• Deploy CLI Overview - All deploy commands
• API Deploy - Deploy as local API server
• Docker Deploy - Deploy to Docker
• AWS Deploy - Deploy to AWS ECS/Fargate