deploy-model
Deploy Azure OpenAI Models
Deploy OpenAI models to Azure Foundry with intelligent routing. This skill handles preset deployments, custom configurations, and capacity discovery across Azure regions.
Download the skill ZIP
Upload in Claude
Go to Settings → Capabilities → Skills → Upload skill
Toggle on and start using
Test it
Using "deploy-model". Deploy gpt-4o to Azure Foundry
Expected outcome:
- Deployment created successfully
- Model: gpt-4o
- Region: eastus
- SKU: Standard
- Capacity: Auto-scaled
- Deployment URL: https://portal.azure.com/...
Using "deploy-model". Find capacity for gpt-4o-mini
Expected outcome:
- Available regions found:
- 1. eastus - 500 TPM available
- 2. westus2 - 300 TPM available
- 3. uksouth - 200 TPM available
Security Audit
Low RiskStatic analysis flagged 475 potential issues, but evaluation confirms these are false positives or expected behavior. The skill uses shell commands to invoke Azure CLI for model deployment operations, which is legitimate for Azure tooling. Network access is to Azure API endpoints. Filesystem patterns are documentation examples, not vulnerabilities. This is an official Microsoft skill for Azure Foundry.
Low Risk Issues (3)
Risk Factors
⚡ Contains scripts
🌐 Network access (2)
Quality Score
What You Can Build
Quick Model Deployment
Deploy an OpenAI model with default settings for rapid prototyping
Custom Enterprise Deployment
Configure deployment with specific SKU, capacity, and RAI policies for production
Capacity Discovery
Find available regions and capacity for specific model types before deployment
Try These Prompts
Deploy gpt-4o to Azure Foundry in the eastus region
Deploy gpt-4 with Standard SKU, 100 TPM capacity, and content filtering enabled to westus2
Find where I can deploy gpt-4o-mini with at least 50 TPM capacity
Show me the best regions for deploying gpt-4 with high capacity
Best Practices
- Always verify capacity availability before deploying in production
- Use descriptive deployment names for easier management
- Configure RAI policies appropriate for your use case
- Store deployment URLs securely for reference
Avoid
- Do not use for listing existing deployments (use foundry_models_deployments_list)
- Do not use for deleting resources (use dedicated Azure tools)
- Do not assume all regions have equal capacity
- Avoid deploying without checking availability first