Explanation:

To deploy an agent and connect it to an existing Azure AI Search index (which already contains vectors), you first need the connection ID for the search resource. Using the Azure AI Projects/Agents client, you obtain that connection ID from your project's connections. Next, you configure the Azure AI Search tool by specifying the index_connection_id, index_name, and (optionally) the query type (simple, semantic, vector, or hybrid), which supports vectorized indexes. After the tool is configured, you create the agent with the tool enabled (passing tools and tool_resources). Finally, you validate the integration by starting a thread and asking questions grounded in the Azure AI Search index. This sequence matches the Microsoft guidance and minimizes effort because it reuses the existing index and only adds a configured tool to a new agent.
Microsoft References
* Use an existing index with the Azure AI Search tool (sequence: create client # configure tool # create agent with tool # ask questions). Microsoft Learn
* Azure AI Search tool overview (add the tool to an agent; supported search types including vector and hybrid). Microsoft Learn
* AzureAISearchTool and AzureAISearchQueryType (parameters, including connection ID and query types: simple, semantic, vector, hybrid).
* Vector and hybrid search in Azure AI Search (background for vector-enabled indexes and hybrid queries).