Azure AI Foundry MCP

Example Use Cases

• 1

  Discover and compare available models (catalog, Labs, model cards, quotas) to select the best model for a task.

• 2

  Index, update, and query documents via Azure AI Search (create/modify/delete indexes, manage documents and indexers).

• 3

  Run evaluations or finetuning workflows and manage related artifacts, events, and metrics for model testing and deployment.

Description

This MCP server exposes a broad set of tools to discover and prototype models from the Azure AI Foundry catalog and Labs, manage Azure AI Search indexes and documents, run text and agent evaluations, and interact with fine-tuning and deployment APIs. It is intended to simplify prototyping and integration workflows by providing discover, build, deploy, knowledge, and evaluation capabilities in a single interface. The project is provided under an MIT license and includes quick-start instructions for VS Code and environment configuration examples.