Skip to main content
Once your provider passes local tests, deploy it to make your resource types available to users.

​
Prerequisites

Before deploying, ensure:
  • Your provider has been tested locally with ProviderHarness
  • You are authenticated: pragma auth login
  • You are in your provider project directory (contains pyproject.toml)

​
Build and Deploy

Build and deploy your provider in a single command:
This:
  1. Uploads your provider code
  2. Builds the provider
  3. Extracts and registers resource type schemas
  4. Deploys the provider
Example output:

​
Stream Build Logs

To see build output in real-time:

​
Build Without Deploying

Build only (to verify the build succeeds before deploying):
Then deploy when ready:

​
Deploy a Specific Version

Deploy a specific version (useful for rollbacks):
List available versions:

​
Check Deployment Status

Example output:

​
Verify Deployment

After deployment, verify your resources are available:
Create a test resource to confirm everything works:

​
Troubleshooting

​
”Could not detect provider package”

Run commands from your provider directory, or specify the package explicitly:

​
“Authentication required”

Log in to the platform:

​
Build failed

Check build logs for details:
Common causes:
  • Missing dependencies in pyproject.toml
  • Syntax errors in provider code
  • Invalid resource schemas

​
Resource type not found

Ensure your resources are properly decorated:
Resources must be exported from the package’s __init__.py for discovery.

​
Deployment not handling events

Verify the provider is running:
If resources remain in PENDING state, check:
  • Provider deployed successfully
  • Resource configuration is valid
  • Dependencies are in READY state

​
Complete Deployment Workflow

Here is the typical workflow for deploying provider changes:
For iterative development:

​
Next Steps

Lifecycle Methods

Implement create, update, and delete operations.

CLI Reference

Full command reference for provider management.