Introduction
Cost management is an essential aspect of designing Azure architectures. However, it is often relegated to a secondary step after technical decisions, which can create unnecessary interruptions in the process. The Azure Pricing Skill for GitHub Copilot aims to solve this problem by integrating cost estimates directly into the workflow, thanks to the Azure MCP tool.
Why this solution is crucial over time
The complexity of cost estimates increases with the size and diversity of environments to manage. Here are some cases where this can be problematic:
- Multi-service architecture distributed across multiple regions.
- Cost calculations during platform reviews or multi-team deployments.
- Collection of reliable costs to help make real-time decisions.
With the integration of the Azure Pricing Skill in GitHub Copilot, costs become a key dimension when choosing a specific Azure service, a region, or a deployment method.
Good to know
This integration helps reduce confusion caused by inaccurate preliminary estimates and improves confidence in the data.
Main features of the Azure Pricing Skill
The Azure Pricing Skill uses the Azure MCP tool via the ms-azuretools.vscode-azure-github-copilot extension within VS Code. Here's how it works:
- Accepted parameters: The MCP can handle parameters such as SKU, service, region, pricing type, currency, and OData filters.
- Structured data rendering: In addition to raw prices returned by the Azure Retail Prices API, the skill provides monthly/annual estimates and reservation option comparisons.
- Local optimization: By default, the skill uses GBP currency to avoid any misunderstanding when discussing in a British context.
Install the VS Code extension
Download and install the ms-azuretools.vscode-azure-github-copilot extension to access the integrated MCP tool.
Enable the MCP
Enable the Azure MCP tool in your VS Code environment. No additional configuration of the mcp.json file is necessary.
1az extension add --name azure-mcpRun Copilot commands
Query GitHub Copilot with simple requests to get structured cost estimates in real-time.
1/azure-pricing How much does a Standard_D4s_v5 cost in UK South?Example of generated results
A typical example is finding the cost of a specific SKU:
Request:
1/azure-pricing How much does a Standard_D4s_v5 cost in UK South?Result:
| Option | Price | Savings compared to PAYG |
|---|---|---|
| Pay-as-you-go | £0.164/hr | — |
| Spot | £0.0213/hr | ~87% |
| 1-Year Reservation | ~£0.097/hr (£847 total) | ~41% |
| 3-Year Reservation | ~£0.061/hr (£1,595 total) | ~63% |
Advantages of this approach
- Automation in Infrastructure-as-Code (IaC): You can ask GitHub Copilot to calculate costs defined in a Terraform or Bicep file.
- Increased precision: Tools ask for clarifications on consumption-based services, such as Azure Functions or Cosmos DB.
- Regional comparisons: Identify cost differences across multiple regions and assess their relevance against latency or operational preferences.
Tip
Integrate this skill into your architecture discussions and code reviews to avoid costly mistakes.
Conclusion
With the integration of the Azure Pricing Skill in GitHub Copilot, costs shift from delayed validation to a central parameter of technical thinking. This offers several benefits:
- Fluidity in engineering decisions: Compare costs from the initial choice of services.
- Centralization of knowledge: Calculation logic documented in the
references/COST-FORMULAS.mdfile enables standardization across teams.
IT professionals and platform teams will find here a powerful tool to establish rational workflows and reduce interruptions in their technical discussions.
To test or contribute to this skill, find the project on the github-copilot-agent-skills repo.
