foundry-agent-sync

To register and synchronize AI agents in Azure AI Foundry, use this Skill to process a local JSON manifest. It applies an idempotent workflow via REST API calls to create or update agent definitions, making them immediately invocable.

Manifest-based deployment for Foundry agents is the process of defining agent configurations in a local JSON file and syncing them directly to Azure AI Foundry. This reconciles definitions without duplicating agents.

Yes, you can update existing Azure AI Foundry agents without duplicating them. The sync workflow is idempotent, safely reconciling agent definitions and metadata from your JSON manifest with existing Foundry records.

To make AI agents invocable through the Foundry portal or REST API, run the sync-foundry-agents script with your local manifest path. This registers the agent definitions, rendering them immediately available for invocation.

The best way to deploy multiple prompt-based agents to Azure AI Foundry is using a manifest-driven approach. Storing agent definitions in a JSON repository file allows automated, idempotent synchronization via REST API calls.

No, you do not need to manually configure agent metadata in the Azure portal after syncing. The manifest-based payload automatically manages agent definitions and metadata, ensuring they are fully updated and immediately invocable.