@@ -125,7 +125,7 @@ To configure access for this solution:
 
    + `AZURE_OPENAI_ENDPOINT` is on the **Endpoints** page of your project's parent resource.
 
-1. For keyless authentication with Microsoft Entra ID, sign in to your Azure account. If you have multiple subscriptions, select the one that contains your Azure AI Search service and Microsoft Foundry project.
+1. For keyless authentication with Microsoft Entra ID, sign in to your Azure account. If you have multiple subscriptions, select the one that contains your Azure AI Search and Microsoft Foundry resources.
 
     ```azurecli
     az login

@@ -8,7 +8,7 @@ ms.service: azure-ai-search
 ms.custom:
   - build-2025
 ms.topic: reference
-ms.date: 02/12/2026
+ms.date: 02/26/2026
 ---
 
 # GenAI Prompt skill
@@ -24,7 +24,7 @@ Here are some examples of how the GenAI prompt skill can help you create content
 - Simplify complex content
 - Perform any other task that you can articulate in a prompt
 
-The GenAI Prompt skill is available in the [latest preview REST API](/rest/api/searchservice/skillsets/create?view=rest-searchservice-2025-11-01-preview&preserve-view=true). This skill supports text, image, and multimodal content, such as a PDF that contains text and images.
+The GenAI Prompt skill is available in the [latest preview REST API](/rest/api/searchservice/skillsets/create?view=rest-searchservice-2025-11-01-preview&preserve-view=true). This skill supports text, image, and multimodal content, such as images with visuals and text extracted from PDF files.
 
 > [!TIP]
 > It's common to use this skill combined with a data chunking skill. The following tutorials demonstrate image verbalization with two different data chunking techniques:

@@ -9,7 +9,7 @@ ms.service: azure-ai-search
 ms.custom:
   - ignite-2025
 ms.topic: quickstart
-ms.date: 02/05/2026
+ms.date: 02/23/2026
 ---
 
 # Quickstart: Agentic retrieval in the Azure portal
@@ -39,9 +39,9 @@ Afterwards, you test the knowledge base by submitting a complex query that requi
 
 + A [Microsoft Foundry project](/azure/ai-foundry/how-to/create-projects) and resource. When you create a project, the resource is automatically created.
 
-+ For text-to-vector conversion, an embedding model [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai). You can use any `text-embedding` model, such as `text-embedding-3-large`.
++ An embedding model [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai) for text-to-vector conversion. You can use any `text-embedding` model, such as `text-embedding-3-large`.
 
-+ For query planning and answer generation, an LLM [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai). You can use any [portal-supported LLM](#supported-llms).
++ An LLM [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai) for query planning and answer generation. You can use any [portal-supported LLM](#supported-llms).
 
 ### Supported LLMs
 
@@ -55,45 +55,23 @@ Although agentic retrieval [programmatically supports several LLMs](agentic-retr
 
 ## Configure access
 
-Before you begin, make sure you have permissions to access content and operations. We recommend Microsoft Entra ID for authentication and role-based access for authorization. You must be an **Owner** or **User Access Administrator** to assign roles. If roles aren't feasible, use [key-based authentication](search-security-api-keys.md) instead.
+Before you begin, make sure you have permissions to access content and operations. This quickstart uses Microsoft Entra ID for authentication and role-based access for authorization. You must be an **Owner** or **User Access Administrator** to assign roles. If roles aren't feasible, use [key-based authentication](search-security-api-keys.md) instead.
 
-To configure access for this quickstart, select each of the following tabs.
+To configure access for this quickstart:
 
-### [Azure AI Search](#tab/search-perms)
+1. Sign in to the [Azure portal](https://portal.azure.com).
 
-Azure AI Search provides the agentic retrieval pipeline. Configure access for yourself and your search service to read and write data, interact with other Azure services, and run the pipeline.
+1. On your Azure AI Search service:
 
-On your Azure AI Search service:
+    1. [Enable role-based access](search-security-enable-roles.md).
 
-1. [Enable role-based access](search-security-enable-roles.md).
+    1. [Create a system-assigned managed identity](search-how-to-managed-identities.md#create-a-system-managed-identity).
 
-1. [Configure a system-assigned managed identity](search-how-to-managed-identities.md#create-a-system-managed-identity).
+    1. [Assign the following roles](search-security-rbac.md) to your user account: **Search Service Contributor**, **Search Index Data Contributor**, and **Search Index Data Reader**.
 
-1. [Assign the following roles](search-security-rbac.md) to yourself.
+1. On your Azure Blob Storage account, assign **Storage Blob Data Contributor** to the managed identity of your search service.
 
-   + **Search Service Contributor**
-
-   + **Search Index Data Contributor**
-
-   + **Search Index Data Reader**
-
-### [Azure Blob Storage](#tab/storage-perms)
-
-Azure Blob Storage stores your source documents for indexing and retrieval. Grant your search service permission to read, write, and delete the documents.
-
-On your Azure Blob Storage account:
-
-+ Assign **Storage Blob Data Reader** to your [search service identity](search-how-to-managed-identities.md#create-a-system-managed-identity).
-
-### [Microsoft Foundry](#tab/foundry-perms)
-
-Microsoft Foundry provides the Azure OpenAI models used for embeddings, query planning, and answer generation. Grant your search service permission to use these models.
-
-On your Microsoft Foundry resource:
-
-+ Assign **Cognitive Services User** to your [search service identity](search-how-to-managed-identities.md#create-a-system-managed-identity).
-
----
+1. On your Microsoft Foundry resource, assign **Cognitive Services User** to the managed identity of your search service.
 
 > [!IMPORTANT]
 > Agentic retrieval has two token-based billing models:

@@ -4,7 +4,7 @@ author: haileytap
 ms.author: haileytapia
 ms.service: azure-ai-search
 ms.topic: include
-ms.date: 01/14/2026
+ms.date: 02/23/2026
 ms.custom: dev-focus
 ai-usage: ai-assisted
 ---
@@ -15,10 +15,10 @@ In this quickstart, you use [agentic retrieval](../../agentic-retrieval-overview
 
 A *knowledge base* orchestrates agentic retrieval by decomposing complex queries into subqueries, running the subqueries against one or more *knowledge sources*, and returning results with metadata. By default, the knowledge base outputs raw content from your sources, but this quickstart uses the answer synthesis output mode for natural-language answer generation.
 
-Although you can provide your own data, this quickstart uses [sample JSON documents](https://github.com/Azure-Samples/azure-search-sample-data/tree/main/nasa-e-book/earth-at-night-json) from NASA's Earth at Night e-book. The documents describe general science topics and images of Earth at night as observed from space.
+Although you can use your own data, this quickstart uses [sample JSON documents](https://github.com/Azure-Samples/azure-search-sample-data/tree/main/nasa-e-book/earth-at-night-json) from NASA's Earth at Night e-book.
 
 > [!TIP]
-> Want to get started right away? See the [azure-search-dotnet-samples](https://github.com/Azure-Samples/azure-search-dotnet-samples/tree/main/quickstart-agentic-retrieval) GitHub repository.
+> Want to get started right away? Download the [source code](https://github.com/Azure-Samples/azure-search-dotnet-samples/tree/main/quickstart-agentic-retrieval) on GitHub.
 
 ## Prerequisites
 
@@ -28,372 +28,64 @@ Although you can provide your own data, this quickstart uses [sample JSON docume
 
 + A [Microsoft Foundry project](/azure/ai-foundry/how-to/create-projects) and resource. When you create a project, the resource is automatically created.
 
-+ The [Azure CLI](/cli/azure/install-azure-cli) for keyless authentication with Microsoft Entra ID.
-
-+ [Visual Studio Code](https://code.visualstudio.com/download).
++ An embedding model [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai) for text-to-vector conversion. You can use any `text-embedding` model, such as `text-embedding-3-large`.
 
-[!INCLUDE [Setup](./agentic-retrieval-setup.md)]
++ An LLM [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai) for query planning and answer generation. You can use any [supported LLM](../../agentic-retrieval-how-to-create-knowledge-base.md#supported-models), such as `gpt-5-mini`.
 
-## Set up the environment
++ [.NET 8](https://dotnet.microsoft.com/download/dotnet/8.0) or later.
 
-To set up the console application for this quickstart:
++ [Visual Studio Code](https://code.visualstudio.com/download).
 
-1. Create a folder named `quickstart-agentic-retrieval` to contain the application.
++ [Git](https://git-scm.com/downloads) to clone the sample repository.
 
-1. Open the folder in Visual Studio Code.
++ The [Azure CLI](/cli/azure/install-azure-cli) for keyless authentication with Microsoft Entra ID.
 
-1. Select **Terminal** > **New Terminal**, and then run the following command to create a console application.
+[!INCLUDE [agentic retrieval setup](agentic-retrieval-setup.md)]
 
-    ```console
-    dotnet new console
-    ```
+## Set up the environment
 
-1. Install the [Azure AI Search client library for .NET](/dotnet/api/overview/azure/search.documents-readme).
+1. Use Git to clone the sample repository.
 
-    ```console
-    dotnet add package Azure.Search.Documents --version 11.8.0-beta.1
+    ```bash
+    git clone https://github.com/Azure-Samples/azure-search-dotnet-samples
     ```
 
-1. Install the `dotenv.net` package to load environment variables from a `.env` file.
+1. Navigate to the quickstart folder and open it in Visual Studio Code.
 
-    ```console
-    dotnet add package dotenv.net
+    ```bash
+    cd azure-search-dotnet-samples/quickstart-agentic-retrieval
+    code .
     ```
 
-1. For keyless authentication with Microsoft Entra ID, install the [Azure.Identity](https://www.nuget.org/packages/Azure.Identity) package.
-
-    ```console
-    dotnet add package Azure.Identity
-    ```
+1. In `sample.env`, replace the placeholder values for `SEARCH_ENDPOINT` and `AOAI_ENDPOINT` with the URLs you obtained in [Get endpoints](#get-endpoints).
 
-1. For keyless authentication with Microsoft Entra ID, sign in to your Azure account. If you have multiple subscriptions, select the one that contains your Azure AI Search service and Microsoft Foundry project.
+1. Rename `sample.env` to `.env`.
 
-    ```console
-    az login
+    ```bash
+    mv sample.env .env
     ```
 
-## Run the code
+1. Install the dependencies.
 
-To create and run the agentic retrieval pipeline:
+    ```bash
+    dotnet restore AgenticRetrievalQuickstart.csproj
+    ```
 
-1. Create a file named `.env` in the `quickstart-agentic-retrieval` folder.
+    When the restore completes, verify that no errors appear in the output.
 
-1. Paste the following environment variables into the `.env` file.
+1. For keyless authentication with Microsoft Entra ID, sign in to your Azure account. If you have multiple subscriptions, select the one that contains your Azure AI Search and Microsoft Foundry resources.
 
-    ```
-    SEARCH_ENDPOINT = PUT-YOUR-SEARCH-SERVICE-URL-HERE
-    AOAI_ENDPOINT = PUT-YOUR-AOAI-FOUNDRY-URL-HERE
+    ```bash
+    az login
     ```
 
-1. Set `SEARCH_ENDPOINT` and `AOAI_ENDPOINT` to the values you obtained in [Get endpoints](#get-endpoints).
-
-1. Paste the following code into the `Program.cs` file.
-
-    ```csharp
-    using dotenv.net;
-    using System.Text.Json;
-    using Azure.Identity;
-    using Azure.Search.Documents;
-    using Azure.Search.Documents.Indexes;
-    using Azure.Search.Documents.Indexes.Models;
-    using Azure.Search.Documents.KnowledgeBases;
-    using Azure.Search.Documents.KnowledgeBases.Models;
-    
-    namespace AzureSearch.Quickstart
-    {
-        class Program
-        {
-            static async Task Main(string[] args)
-            {
-                // Load environment variables from the .env file
-                // Ensure your .env file is in the same directory with the required variables
-                DotEnv.Load();
-    
-                string searchEndpoint = Environment.GetEnvironmentVariable("SEARCH_ENDPOINT")
-                    ?? throw new InvalidOperationException("SEARCH_ENDPOINT isn't set.");
-                string aoaiEndpoint = Environment.GetEnvironmentVariable("AOAI_ENDPOINT")
-                    ?? throw new InvalidOperationException("AOAI_ENDPOINT isn't set.");
-    
-                string aoaiEmbeddingModel = "text-embedding-3-large";
-                string aoaiEmbeddingDeployment = "text-embedding-3-large";
-                string aoaiGptModel = "gpt-5-mini";
-                string aoaiGptDeployment = "gpt-5-mini";
-    
-                string indexName = "earth-at-night";
-                string knowledgeSourceName = "earth-knowledge-source";
-                string knowledgeBaseName = "earth-knowledge-base";
-    
-                var credential = new DefaultAzureCredential();
-    
-                // Define fields for the index
-                var fields = new List<SearchField>
-                {
-                    new SimpleField("id", SearchFieldDataType.String) { IsKey = true, IsFilterable = true, IsSortable = true, IsFacetable = true },
-                    new SearchField("page_chunk", SearchFieldDataType.String) { IsFilterable = false, IsSortable = false, IsFacetable = false },
-                    new SearchField("page_embedding_text_3_large", SearchFieldDataType.Collection(SearchFieldDataType.Single)) { VectorSearchDimensions = 3072, VectorSearchProfileName = "hnsw_text_3_large" },
-                    new SimpleField("page_number", SearchFieldDataType.Int32) { IsFilterable = true, IsSortable = true, IsFacetable = true }
-                };
-    
-                // Define a vectorizer
-                var vectorizer = new AzureOpenAIVectorizer(vectorizerName: "azure_openai_text_3_large")
-                {
-                    Parameters = new AzureOpenAIVectorizerParameters
-                    {
-                        ResourceUri = new Uri(aoaiEndpoint),
-                        DeploymentName = aoaiEmbeddingDeployment,
-                        ModelName = aoaiEmbeddingModel
-                    }
-                };
-    
-                // Define a vector search profile and algorithm
-                var vectorSearch = new VectorSearch()
-                {
-                    Profiles =
-                    {
-                        new VectorSearchProfile(
-                            name: "hnsw_text_3_large",
-                            algorithmConfigurationName: "alg"
-                        )
-                        {
-                            VectorizerName = "azure_openai_text_3_large"
-                        }
-                    },
-                    Algorithms =
-                    {
-                        new HnswAlgorithmConfiguration(name: "alg")
-                    },
-                    Vectorizers =
-                    {
-                        vectorizer
-                    }
-                };
-    
-                // Define a semantic configuration
-                var semanticConfig = new SemanticConfiguration(
-                    name: "semantic_config",
-                    prioritizedFields: new SemanticPrioritizedFields
-                    {
-                        ContentFields = { new SemanticField("page_chunk") }
-                    }
-                );
-    
-                var semanticSearch = new SemanticSearch()
-                {
-                    DefaultConfigurationName = "semantic_config",
-                    Configurations = { semanticConfig }
-                };
-    
-                // Create the index
-                var index = new SearchIndex(indexName)
-                {
-                    Fields = fields,
-                    VectorSearch = vectorSearch,
-                    SemanticSearch = semanticSearch
-                };
-    
-                // Create the index client, deleting and recreating the index if it exists
-                var indexClient = new SearchIndexClient(new Uri(searchEndpoint), credential);
-                await indexClient.CreateOrUpdateIndexAsync(index);
-                Console.WriteLine($"Index '{indexName}' created or updated successfully.");
-    
-                // Upload sample documents from the GitHub URL
-                string url = "https://raw.githubusercontent.com/Azure-Samples/azure-search-sample-data/refs/heads/main/nasa-e-book/earth-at-night-json/documents.json";
-                var httpClient = new HttpClient();
-                var response = await httpClient.GetAsync(url);
-                response.EnsureSuccessStatusCode();
-                var json = await response.Content.ReadAsStringAsync();
-                var documents = JsonSerializer.Deserialize<List<Dictionary<string, object>>>(json);
-                var searchClient = new SearchClient(new Uri(searchEndpoint), indexName, credential);
-                var searchIndexingBufferedSender = new SearchIndexingBufferedSender<Dictionary<string, object>>(
-                    searchClient,
-                    new SearchIndexingBufferedSenderOptions<Dictionary<string, object>>
-                    {
-                        KeyFieldAccessor = doc => doc["id"].ToString(),
-                    }
-                );
-
-                await searchIndexingBufferedSender.UploadDocumentsAsync(documents);
-                await searchIndexingBufferedSender.FlushAsync();
-                Console.WriteLine($"Documents uploaded to index '{indexName}' successfully.");
-    
-                // Create a knowledge source
-                var indexKnowledgeSource = new SearchIndexKnowledgeSource(
-                    name: knowledgeSourceName,
-                    searchIndexParameters: new SearchIndexKnowledgeSourceParameters(searchIndexName: indexName)
-                    {
-                        SourceDataFields = { new SearchIndexFieldReference(name: "id"), new SearchIndexFieldReference(name: "page_chunk"), new SearchIndexFieldReference(name: "page_number") }
-                    }
-                );
-    
-                await indexClient.CreateOrUpdateKnowledgeSourceAsync(indexKnowledgeSource);
-                Console.WriteLine($"Knowledge source '{knowledgeSourceName}' created or updated successfully.");
-    
-                // Create a knowledge base
-                var openAiParameters = new AzureOpenAIVectorizerParameters
-                {
-                    ResourceUri = new Uri(aoaiEndpoint),
-                    DeploymentName = aoaiGptDeployment,
-                    ModelName = aoaiGptModel
-                };
-    
-                var model = new KnowledgeBaseAzureOpenAIModel(azureOpenAIParameters: openAiParameters);
-    
-                var knowledgeBase = new KnowledgeBase(
-                    name: knowledgeBaseName,
-                    knowledgeSources: new KnowledgeSourceReference[] { new KnowledgeSourceReference(knowledgeSourceName) }
-                )
-                {
-                    RetrievalReasoningEffort = new KnowledgeRetrievalLowReasoningEffort(),
-                    AnswerInstructions = "Provide a two sentence concise and informative answer based on the retrieved documents.",
-                    Models = { model }
-                };
-    
-                await indexClient.CreateOrUpdateKnowledgeBaseAsync(knowledgeBase);
-                Console.WriteLine($"Knowledge base '{knowledgeBaseName}' created or updated successfully.");
-    
-                // Set up messages
-                string instructions = @"A Q&A agent that can answer questions about the Earth at night.
-                If you don't have the answer, respond with ""I don't know"".";
-
-                var messages = new List<Dictionary<string, string>>
-                {
-                    new Dictionary<string, string>
-                    {
-                        { "role", "system" },
-                        { "content", instructions }
-                    }
-                };
-    
-                // Run agentic retrieval
-                var baseClient = new KnowledgeBaseRetrievalClient(
-                    endpoint: new Uri(searchEndpoint),
-                    knowledgeBaseName: knowledgeBaseName,
-                    tokenCredential: new DefaultAzureCredential()
-                );
-
-                string query = @"Why do suburban belts display larger December brightening than urban cores even though absolute light levels are higher downtown? Why is the Phoenix nighttime street grid is so sharply visible from space, whereas large stretches of the interstate between midwestern cities remain comparatively dim?";
-
-                messages.Add(new Dictionary<string, string>
-                {
-                    { "role", "user" },
-                    { "content", query }
-                });
-
-                Console.WriteLine($"Running the query...{query}");
-                var retrievalRequest = new KnowledgeBaseRetrievalRequest();
-                foreach (Dictionary<string, string> message in messages) {
-                    if (message["role"] != "system") {
-                        retrievalRequest.Messages.Add(new KnowledgeBaseMessage(content: new[] { new KnowledgeBaseMessageTextContent(message["content"]) }) { Role = message["role"] });
-                    }
-                }
-                retrievalRequest.RetrievalReasoningEffort = new KnowledgeRetrievalLowReasoningEffort();
-                var retrievalResult = await baseClient.RetrieveAsync(retrievalRequest);
-    
-                messages.Add(new Dictionary<string, string>
-                {
-                    { "role", "assistant" },
-                    { "content", (retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent)!.Text }
-                });
-    
-                // Print the response, activity, and references
-                Console.WriteLine("Response:");
-                Console.WriteLine((retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent)!.Text);
-    
-                Console.WriteLine("Activity:");
-                foreach (var activity in retrievalResult.Value.Activity)
-                {
-                    Console.WriteLine($"Activity Type: {activity.GetType().Name}");
-                    string activityJson = JsonSerializer.Serialize(
-                        activity,
-                        activity.GetType(),
-                        new JsonSerializerOptions { WriteIndented = true }
-                    );
-                    Console.WriteLine(activityJson);
-                }
-    
-                Console.WriteLine("References:");
-                foreach (var reference in retrievalResult.Value.References)
-                {
-                    Console.WriteLine($"Reference Type: {reference.GetType().Name}");
-                    string referenceJson = JsonSerializer.Serialize(
-                        reference,
-                        reference.GetType(),
-                        new JsonSerializerOptions { WriteIndented = true }
-                    );
-                    Console.WriteLine(referenceJson);
-                }
-    
-                // Continue the conversation
-                string nextQuery = "How do I find lava at night?";
-                Console.WriteLine($"Continue the conversation with this query: {nextQuery}");
-                messages.Add(new Dictionary<string, string>
-                {
-                    { "role", "user" },
-                    { "content", nextQuery }
-                });
-    
-                retrievalRequest = new KnowledgeBaseRetrievalRequest();
-                foreach (Dictionary<string, string> message in messages) {
-                    if (message["role"] != "system") {
-                        retrievalRequest.Messages.Add(new KnowledgeBaseMessage(content: new[] { new KnowledgeBaseMessageTextContent(message["content"]) }) { Role = message["role"] });
-                    }
-                }
-                retrievalRequest.RetrievalReasoningEffort = new KnowledgeRetrievalLowReasoningEffort();
-                retrievalResult = await baseClient.RetrieveAsync(retrievalRequest);
-    
-                messages.Add(new Dictionary<string, string>
-                {
-                    { "role", "assistant" },
-                    { "content", (retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent)!.Text }
-                });
-    
-                // Print the new response, activity, and references
-                Console.WriteLine("Response:");
-                Console.WriteLine((retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent)!.Text);
-    
-                Console.WriteLine("Activity:");
-                foreach (var activity in retrievalResult.Value.Activity)
-                {
-                    Console.WriteLine($"Activity Type: {activity.GetType().Name}");
-                    string activityJson = JsonSerializer.Serialize(
-                        activity,
-                        activity.GetType(),
-                        new JsonSerializerOptions { WriteIndented = true }
-                    );
-                    Console.WriteLine(activityJson);
-                }
-    
-                Console.WriteLine("References:");
-                foreach (var reference in retrievalResult.Value.References)
-                {
-                    Console.WriteLine($"Reference Type: {reference.GetType().Name}");
-                    string referenceJson = JsonSerializer.Serialize(
-                        reference,
-                        reference.GetType(),
-                        new JsonSerializerOptions { WriteIndented = true }
-                    );
-                    Console.WriteLine(referenceJson);
-                }
-    
-                // Clean up resources
-                await indexClient.DeleteKnowledgeBaseAsync(knowledgeBaseName);
-                Console.WriteLine($"Knowledge base '{knowledgeBaseName}' deleted successfully.");
-    
-                await indexClient.DeleteKnowledgeSourceAsync(knowledgeSourceName);
-                Console.WriteLine($"Knowledge source '{knowledgeSourceName}' deleted successfully.");
-    
-                await indexClient.DeleteIndexAsync(indexName);
-                Console.WriteLine($"Index '{indexName}' deleted successfully.");
-            }
-        }
-    }
-    ```
+## Run the code
 
-1. Build and run the application.
+Run the application to create an index, upload documents, configure a knowledge source and knowledge base, and run agentic retrieval queries.
 
-    ```console
-    dotnet run
-    ```
+```bash
+dotnet run --project AgenticRetrievalQuickstart.csproj
+```
 
 ### Output
 
@@ -404,91 +96,56 @@ Index 'earth-at-night' created or updated successfully.
 Documents uploaded to index 'earth-at-night' successfully.
 Knowledge source 'earth-knowledge-source' created or updated successfully.
 Knowledge base 'earth-knowledge-base' created or updated successfully.
+Running the query...Why do suburban belts display larger December brightening than urban cores even though absolute light levels are higher downtown? Why is the Phoenix nighttime street grid is so sharply visible from space, whereas large stretches of the interstate between midwestern cities remain comparatively dim?
 Response:
-Suburban belts show larger December brightening because holiday displays concentrate in suburbs and outskirts where there is more yard space and many single‑family homes [ref_id:5], while urban cores—already having higher absolute light levels—tend to show smaller relative increases (central areas typically brighten ~20–30%) [ref_id:8][ref_id:5]. Phoenix’s nighttime street grid is sharply visible because the metropolitan area is laid out on a regular, continuously lit grid with bright commercial and industrial nodes along major corridors like Grand Avenue [ref_id:0][ref_id:3], whereas long interstate stretches between Midwestern cities cross sparsely populated or rural regions with far fewer continuous roadside lights and so appear comparatively dim [ref_id:8].
+Suburban belts brighten more (in relative terms) in December because holiday lighting is concentrated in yards and single-family suburbs where extra decorative lights add a large percentage increase over baseline residential lighting, whereas dense urban cores already have high absolute light levels so the same added holiday lights produce a smaller relative change [ref_id:6][ref_id:2]. The documents note suburbs and outskirts show the biggest holiday increases and central urban areas show smaller percent increases (20-30% in cores, larger in suburbs) linked to available yard space and prevalence of single-family homes [ref_id:6][ref_id:2]. Phoenix's street grid appears very sharp from space because the city's regular, closely spaced north-south/east-west street lighting and bright nodes at intersections and commercial strips produce a strong, high-contrast patterned signal (grid and major corridors like Grand Avenue), while long Midwestern interstates are comparatively dim because roadway lighting is more sparse and continuous between cities and navigable rivers/long highways lack the dense, closely spaced light sources that produce visible nodes and grid patterns at night [ref_id:3][ref_id:7]. In addition, the Black Marble processing accounts for atmospheric, lunar, snow/seasonal and stray-light effects and isolates artificial emissions, so concentrated urban lighting (as in Phoenix) stands out more in the corrected radiance product than dispersed or spaced sources like isolated stretches of interstate or sparsely lit rivers and plains [ref_id:1][ref_id:4][ref_id:7].
 Activity:
 Activity Type: KnowledgeBaseModelQueryPlanningActivityRecord
 {
-  "InputTokens": 1350,
-  "OutputTokens": 1314,
+  "InputTokens": 1489,
+  "OutputTokens": 326,
   "Id": 0,
-  "ElapsedMs": 14162,
+  "ElapsedMs": 4558,
   "Error": null
 }
 Activity Type: KnowledgeBaseSearchIndexActivityRecord
 {
   "SearchIndexArguments": {
-    "Search": "Causes of December brightening in satellite nightlights: why suburban belts show larger relative December brightening than urban cores (roles of holiday residential lighting, snow albedo, urban heat island, commercial lighting patterns)",
+    "Search": "December brightening suburban belts vs urban cores light pollution causes seasonal variation \"December brightening\" satellite night lights",
     "Filter": null,
-    "SourceDataFields": [],
+    "SourceDataFields": [
+      {
+        "Name": "page_chunk"
+      },
+      {
+        "Name": "id"
+      },
+      {
+        "Name": "page_number"
+      }
+    ],
     "SearchFields": [],
-    "SemanticConfigurationName": null
+    "SemanticConfigurationName": "semantic_config"
   },
   "KnowledgeSourceName": "earth-knowledge-source",
-  "QueryTime": "2025-11-05T21:56:26.747+00:00",
-  "Count": 19,
+  "QueryTime": "2026-02-24T14:59:41.536+00:00",
+  "Count": 21,
   "Id": 1,
-  "ElapsedMs": 537,
-  "Error": null
-}
-Activity Type: KnowledgeBaseSearchIndexActivityRecord
-{
-  "SearchIndexArguments": {
-    "Search": "Why is Phoenix\u0019s nighttime street grid so sharply visible from space? (effects of streetlight density, luminaire type/aiming, spacing, urban grid layout, traffic vs roadway lighting)",
-    "Filter": null,
-    "SourceDataFields": [],
-    "SearchFields": [],
-    "SemanticConfigurationName": null
-  },
-  "KnowledgeSourceName": "earth-knowledge-source",
-  "QueryTime": "2025-11-05T21:56:27.182+00:00",
-  "Count": 7,
-  "Id": 2,
-  "ElapsedMs": 434,
+  "ElapsedMs": 623,
   "Error": null
 }
-Activity Type: KnowledgeBaseSearchIndexActivityRecord
-{
-  "SearchIndexArguments": {
-    "Search": "How do satellite nightlight sensor characteristics (VIIRS DNB, DMSP-OLS) \u2014 spatial resolution, dynamic range, saturation, blooming \u2014 affect observed brightness and structure of urban cores, suburbs, and long interstate stretches?",
-    "Filter": null,
-    "SourceDataFields": [],
-    "SearchFields": [],
-    "SemanticConfigurationName": null
-  },
-  "KnowledgeSourceName": "earth-knowledge-source",
-  "QueryTime": "2025-11-05T21:56:27.786+00:00",
-  "Count": 23,
-  "Id": 3,
-  "ElapsedMs": 604,
-  "Error": null
-}
-Activity Type: KnowledgeBaseAgenticReasoningActivityRecord
-{
-  "ReasoningTokens": 70232,
-  "RetrievalReasoningEffort": {},
-  "Id": 4,
-  "ElapsedMs": null,
-  "Error": null
-}
-Activity Type: KnowledgeBaseModelAnswerSynthesisActivityRecord
-{
-  "InputTokens": 7467,
-  "OutputTokens": 1710,
-  "Id": 5,
-  "ElapsedMs": 26663,
-  "Error": null
-}
-Results:
+... // Trimmed for brevity
+References:
 Reference Type: KnowledgeBaseSearchIndexReference
 {
-  "DocKey": "earth_at_night_508_page_104_verbalized",
+  "DocKey": "earth_at_night_508_page_105_verbalized",
   "Id": "0",
   "ActivitySource": 2,
   "SourceData": {},
-  "RerankerScore": 2.6344998
+  "RerankerScore": 2.7294974
 }
 ... // Trimmed for brevity
+Continue the conversation with this query: How do I find lava at night?
 Response:
 ... // Trimmed for brevity
 Activity:
@@ -502,6 +159,8 @@ Index 'earth-at-night' deleted successfully.
 
 ## Understand the code
 
+[!INCLUDE [understand code note](../understand-code-note.md)]
+
 Now that you've run the code, let's break down the key steps:
 
 1. [Create a search index](#create-a-search-index)
@@ -734,7 +393,7 @@ var retrievalResult = await baseClient.RetrieveAsync(retrievalRequest);
 messages.Add(new Dictionary<string, string>
 {
     { "role", "assistant" },
-    { "content", (retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent).Text }
+    { "content", (retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent)!.Text }
 });
 ```
 
@@ -753,7 +412,7 @@ The following code displays the response, activity, and references from the retr
 ```csharp
 // Print the response, activity, and references
 Console.WriteLine("Response:");
-Console.WriteLine((retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent).Text);
+Console.WriteLine((retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent)!.Text);
 
 Console.WriteLine("Activity:");
 foreach (var activity in retrievalResult.Value.Activity)
@@ -804,7 +463,7 @@ retrievalResult = await baseClient.RetrieveAsync(retrievalRequest);
 messages.Add(new Dictionary<string, string>
 {
     { "role", "assistant" },
-    { "content", (retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent).Text }
+    { "content", (retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent)!.Text }
 });
 ```
 
@@ -815,7 +474,7 @@ The following code displays the new response, activity, and references from the
 ```csharp
 // Print the new response, activity, and references
 Console.WriteLine("Response:");
-Console.WriteLine((retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent).Text);
+Console.WriteLine((retrievalResult.Value.Response[0].Content[0] as KnowledgeBaseMessageTextContent)!.Text);
 
 Console.WriteLine("Activity:");
 foreach (var activity in retrievalResult.Value.Activity)
@@ -846,7 +505,7 @@ foreach (var reference in retrievalResult.Value.References)
 
 [!INCLUDE [clean up resources (paid)](../resource-cleanup-paid.md)]
 
-Otherwise, the following code from `Program.cs` deleted the objects you created in this quickstart.
+Otherwise, the following code from `program.cs` deleted the objects you created in this quickstart.
 
 ### Delete the knowledge base
 

@@ -4,7 +4,7 @@ author: haileytap
 ms.author: haileytapia
 ms.service: azure-ai-search
 ms.topic: include
-ms.date: 01/14/2026
+ms.date: 02/23/2026
 ms.custom: dev-focus
 ai-usage: ai-assisted
 ---
@@ -15,7 +15,10 @@ In this quickstart, you use [agentic retrieval](../../agentic-retrieval-overview
 
 A *knowledge base* orchestrates agentic retrieval by decomposing complex queries into subqueries, running the subqueries against one or more *knowledge sources*, and returning results with metadata. By default, the knowledge base outputs raw content from your sources, but this quickstart uses the answer synthesis output mode for natural-language answer generation.
 
-Although you can provide your own data, this quickstart uses [sample JSON documents](https://github.com/Azure-Samples/azure-search-sample-data/tree/main/nasa-e-book/earth-at-night-json) from NASA's Earth at Night e-book. The documents describe general science topics and images of Earth at night as observed from space.
+Although you can use your own data, this quickstart uses [sample JSON documents](https://github.com/Azure-Samples/azure-search-sample-data/tree/main/nasa-e-book/earth-at-night-json) from NASA's Earth at Night e-book.
+
+> [!TIP]
+> Source code for the JavaScript version of this quickstart isn't available. You can copy code directly from this article.
 
 ## Prerequisites
 
@@ -25,62 +28,60 @@ Although you can provide your own data, this quickstart uses [sample JSON docume
 
 + A [Microsoft Foundry project](/azure/ai-foundry/how-to/create-projects) and resource. When you create a project, the resource is automatically created.
 
-+ The [Azure CLI](/cli/azure/install-azure-cli) for keyless authentication with Microsoft Entra ID.
++ An embedding model [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai) for text-to-vector conversion. You can use any `text-embedding` model, such as `text-embedding-3-large`.
 
-+ [Visual Studio Code](https://code.visualstudio.com/download) and the latest LTS version of [Node.js](https://nodejs.org/en/download/).
++ An LLM [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai) for query planning and answer generation. You can use any [supported LLM](../../agentic-retrieval-how-to-create-knowledge-base.md#supported-models), such as `gpt-5-mini`.
 
-[!INCLUDE [Setup](./agentic-retrieval-setup.md)]
++ [Node.js 20 LTS](https://nodejs.org/en/download/) or later.
 
-## Set up the environment
++ [Visual Studio Code](https://code.visualstudio.com/download).
+
++ The [Azure CLI](/cli/azure/install-azure-cli) for keyless authentication with Microsoft Entra ID.
 
-To set up the console application for this quickstart:
+[!INCLUDE [agentic retrieval setup](agentic-retrieval-setup.md)]
+
+## Set up the environment
 
 1. Create a folder named `quickstart-agentic-retrieval` to contain the application.
 
 1. Open the folder in Visual Studio Code.
 
-1. Select **Terminal** > **New Terminal**, and then run the following commands to initialize the `package.json` file.
+1. Select **Terminal** > **New Terminal**, and then run the following commands to initialize `package.json`.
 
-    ```console
+    ```bash
     npm init -y
     npm pkg set type=module
     ```
 
 1. Install the [Azure AI Search client library for JavaScript](/javascript/api/overview/azure/search-documents-readme).
 
-    ```console
+    ```bash
     npm install @azure/search-documents@12.3.0-beta.1
     ```
 
 1. For keyless authentication with Microsoft Entra ID, install the [Azure Identity client library for JavaScript](/javascript/api/overview/azure/identity-readme).
 
-    ```console
+    ```bash
     npm install @azure/identity
     ```
 
-1. For keyless authentication with Microsoft Entra ID, sign in to your Azure account. If you have multiple subscriptions, select the one that contains your Azure AI Search service and Microsoft Foundry project.
+1. For keyless authentication with Microsoft Entra ID, sign in to your Azure account. If you have multiple subscriptions, select the one that contains your Azure AI Search and Microsoft Foundry resources.
 
-    ```console
+    ```bash
     az login
     ```
 
 ## Run the code
 
-To create and run the agentic retrieval pipeline:
-
-1. Create a file named `.env` in the `quickstart-agentic-retrieval` folder.
-
-1. Paste the following environment variables into the `.env` file.
+1. Create a file named `.env` in the `quickstart-agentic-retrieval` folder, and then paste the following content. Replace the placeholder values with the URLs you obtained in [Get endpoints](#get-endpoints).
 
     ```
     AZURE_SEARCH_ENDPOINT = https://<your-search-service-name>.search.windows.net
-    AZURE_OPENAI_ENDPOINT = https://<your-ai-foundry-resource-name>.openai.azure.com/
+    AZURE_OPENAI_ENDPOINT = https://<your-foundry-resource-name>.openai.azure.com/
     AZURE_OPENAI_GPT_DEPLOYMENT = gpt-5-mini
     AZURE_OPENAI_EMBEDDING_DEPLOYMENT = text-embedding-3-large
     ```
 
-1. Set `AZURE_SEARCH_ENDPOINT` and `AZURE_OPENAI_ENDPOINT` to the values you obtained in [Get endpoints](#get-endpoints).
-
 1. Create a file named `index.js`, and then paste the following code into the file.
 
     ```javascript
@@ -401,15 +402,15 @@ To create and run the agentic retrieval pipeline:
 
 1. Build and run the application.
 
-    ```console
+    ```bash
     node --env-file=.env index.js
     ```
 
 ### Output
 
 The output of the application should be similar to the following:
 
-```console
+```output
 Waiting for indexing to complete...
 Expected documents: 194
 Current indexed count: 194
@@ -457,58 +458,7 @@ Activity Type: searchIndex
     "semanticConfigurationName": "semantic_config"
   }
 }
-Activity Type: searchIndex
-{
-  "id": 2,
-  "type": "searchIndex",
-  "elapsedMs": 538,
-  "knowledgeSourceName": "earth-knowledge-source",
-  "queryTime": "2025-12-19T15:38:24.001Z",
-  "count": 0,
-  "searchIndexArguments": {
-    "search": "factors that make Phoenix nighttime street grid highly visible from space reasons highway/interstate lighting visibility differences Midwestern interstates dim",
-    "filter": null,
-    "sourceDataFields": [
-      {
-        "name": "page_chunk"
-      },
-      {
-        "name": "id"
-      },
-      {
-        "name": "page_number"
-      }
-    ],
-    "searchFields": [],
-    "semanticConfigurationName": "semantic_config"
-  }
-}
-Activity Type: searchIndex
-{
-  "id": 3,
-  "type": "searchIndex",
-  "elapsedMs": 465,
-  "knowledgeSourceName": "earth-knowledge-source",
-  "queryTime": "2025-12-19T15:38:24.467Z",
-  "count": 2,
-  "searchIndexArguments": {
-    "search": "satellite nighttime lights seasonal variations suburban brightening studies December holiday lighting residential vs commercial lighting patterns",
-    "filter": null,
-    "sourceDataFields": [
-      {
-        "name": "page_chunk"
-      },
-      {
-        "name": "id"
-      },
-      {
-        "name": "page_number"
-      }
-    ],
-    "searchFields": [],
-    "semanticConfigurationName": "semantic_config"
-  }
-}
+... // Trimmed for brevity
 Activity Type: agenticReasoning
 {
   "id": 4,
@@ -570,7 +520,7 @@ Now that you have the code, let's break down the key components:
 1. [Create a knowledge source](#create-a-knowledge-source)
 1. [Create a knowledge base](#create-a-knowledge-base)
 1. [Run the retrieval pipeline](#run-the-retrieval-pipeline)
-1. [Review the response, activity, and references](#review-the-response-activity-and-results)
+1. [Review the response, activity, and references](#review-the-response-activity-and-references)
 1. [Continue the conversation](#continue-the-conversation)
 
 ### Create a search index
@@ -580,7 +530,7 @@ In Azure AI Search, an index is a structured collection of data. The following c
 The index schema contains fields for document identification and page content, embeddings, and numbers. The schema also includes configurations for semantic ranking and vector search, which uses your `text-embedding-3-large` deployment to vectorize text and match documents based on semantic similarity.
 
 ```javascript
-const index: SearchIndex = {
+const index = {
     name: 'earth_at_night',
     fields: [
         {
@@ -590,15 +540,15 @@ const index: SearchIndex = {
             filterable: true,
             sortable: true,
             facetable: true
-        } as SearchField,
+        },
         {
             name: "page_chunk",
             type: "Edm.String",
             searchable: true,
             filterable: false,
             sortable: false,
             facetable: false
-        } as SearchField,
+        },
         {
             name: "page_embedding_text_3_large",
             type: "Collection(Edm.Single)",
@@ -608,54 +558,54 @@ const index: SearchIndex = {
             facetable: false,
             vectorSearchDimensions: 3072,
             vectorSearchProfileName: "hnsw_text_3_large"
-        } as SearchField,
+        },
         {
             name: "page_number",
             type: "Edm.Int32",
             filterable: true,
             sortable: true,
             facetable: true
-        } as SearchField
+        }
     ],
     vectorSearch: {
         profiles: [
             {
                 name: "hnsw_text_3_large",
                 algorithmConfigurationName: "alg",
                 vectorizerName: "azure_openai_text_3_large"
-            } as VectorSearchProfile
+            }
         ],
         algorithms: [
             {
                 name: "alg",
                 kind: "hnsw"
-            } as HnswAlgorithmConfiguration
+            }
         ],
         vectorizers: [
             {
                 vectorizerName: "azure_openai_text_3_large",
                 kind: "azureOpenAI",
                 parameters: {
-                    resourceUrl: process.env.AZURE_OPENAI_ENDPOINT!,
-                    deploymentId: process.env.AZURE_OPENAI_EMBEDDING_DEPLOYMENT!,
-                    modelName: process.env.AZURE_OPENAI_EMBEDDING_DEPLOYMENT!
-                } as AzureOpenAIParameters
-            } as AzureOpenAIVectorizer
+                    resourceUrl: process.env.AZURE_OPENAI_ENDPOINT,
+                    deploymentId: process.env.AZURE_OPENAI_EMBEDDING_DEPLOYMENT,
+                    modelName: process.env.AZURE_OPENAI_EMBEDDING_DEPLOYMENT
+                }
+            }
         ]
-    } as VectorSearch,
+    },
     semanticSearch: {
         defaultConfigurationName: "semantic_config",
         configurations: [
             {
                 name: "semantic_config",
                 prioritizedFields: {
                     contentFields: [
-                        { name: "page_chunk" } as SemanticField
+                        { name: "page_chunk" }
                     ]
-                } as SemanticPrioritizedFields
-            } as SemanticConfiguration
+                }
+            }
         ]
-    } as SemanticSearch
+    }
 };
 
 const credential = new DefaultAzureCredential();
@@ -714,7 +664,7 @@ console.log(`✓ All ${documents.length} documents indexed successfully!`);
 
 A knowledge source is a reusable reference to source data. The following code defines a knowledge source named `earth-knowledge-source` that targets the `earth-at-night` index.
 
-`source_data_fields` specifies which index fields are included in citation references. This example includes only human-readable fields to avoid lengthy, uninterpretable embeddings in responses.
+`sourceDataFields` specifies which index fields are included in citation references. This example includes only human-readable fields to avoid lengthy, uninterpretable embeddings in responses.
 
 ```javascript
 await searchIndexClient.createKnowledgeSource({

@@ -4,7 +4,7 @@ author: haileytap
 ms.author: haileytapia
 ms.service: azure-ai-search
 ms.topic: include
-ms.date: 01/14/2026
+ms.date: 02/23/2026
 ms.custom: dev-focus
 ai-usage: ai-assisted
 ---
@@ -15,10 +15,10 @@ In this quickstart, you use [agentic retrieval](../../agentic-retrieval-overview
 
 A *knowledge base* orchestrates agentic retrieval by decomposing complex queries into subqueries, running the subqueries against one or more *knowledge sources*, and returning results with metadata. By default, the knowledge base outputs raw content from your sources, but this quickstart uses the answer synthesis output mode for natural-language answer generation.
 
-Although you can provide your own data, this quickstart uses [sample JSON documents](https://github.com/Azure-Samples/azure-search-sample-data/tree/main/nasa-e-book/earth-at-night-json) from NASA's Earth at Night e-book. The documents describe general science topics and images of Earth at night as observed from space.
+Although you can use your own data, this quickstart uses [sample JSON documents](https://github.com/Azure-Samples/azure-search-sample-data/tree/main/nasa-e-book/earth-at-night-json) from NASA's Earth at Night e-book.
 
 > [!TIP]
-> Want to get started right away? See the [azure-search-python-samples](https://github.com/Azure-Samples/azure-search-python-samples/tree/main/Quickstart-Agentic-Retrieval) GitHub repository.
+> Want to get started right away? Download the [source code](https://github.com/Azure-Samples/azure-search-python-samples/tree/main/Quickstart-Agentic-Retrieval) on GitHub.
 
 ## Prerequisites
 
@@ -28,327 +28,64 @@ Although you can provide your own data, this quickstart uses [sample JSON docume
 
 + A [Microsoft Foundry project](/azure/ai-foundry/how-to/create-projects) and resource. When you create a project, the resource is automatically created.
 
-+ The [Azure CLI](/cli/azure/install-azure-cli) for keyless authentication with Microsoft Entra ID.
++ An embedding model [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai) for text-to-vector conversion. You can use any `text-embedding` model, such as `text-embedding-3-large`.
 
-+ The latest version of [Python](https://www.python.org/downloads/).
++ An LLM [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai) for query planning and answer generation. You can use any [supported LLM](../../agentic-retrieval-how-to-create-knowledge-base.md#supported-models), such as `gpt-5-mini`.
 
-+ [Visual Studio Code](https://code.visualstudio.com/download) with the [Python extension](https://marketplace.visualstudio.com/items?itemName=ms-python.python).
++ [Python 3.8](https://www.python.org/downloads/) or later.
 
-[!INCLUDE [Setup](./agentic-retrieval-setup.md)]
++ [Visual Studio Code](https://code.visualstudio.com/download) with the [Python](https://marketplace.visualstudio.com/items?itemName=ms-python.python) and [Jupyter](https://marketplace.visualstudio.com/items?itemName=ms-toolsai.jupyter) extensions.
 
-## Connect from your local system
++ [Git](https://git-scm.com/downloads) to clone the sample repository.
 
-You configured role-based access to interact with Azure AI Search and Azure OpenAI in Foundry Models. Use the Azure CLI to sign in to the same subscription and tenant for both resources. For more information, see [Quickstart: Connect without keys](../../search-get-started-rbac.md).
++ The [Azure CLI](/cli/azure/install-azure-cli) for keyless authentication with Microsoft Entra ID.
 
-To connect from your local system:
+[!INCLUDE [agentic retrieval setup](agentic-retrieval-setup.md)]
 
-1. Create a folder named `quickstart-agentic-retrieval`.
+## Set up the environment
 
-1. Open the folder in Visual Studio Code.
+1. Use Git to clone the sample repository.
 
-1. Select **Terminal** > **New Terminal**.
+    ```bash
+    git clone https://github.com/Azure-Samples/azure-search-python-samples
+    ```
 
-1. Run the following command to sign in to your Azure account. If you have multiple subscriptions, select the one that contains your Azure AI Search service and Foundry project.
+1. Navigate to the quickstart folder and open it in Visual Studio Code.
 
-    ```azurecli
-    az login
+    ```bash
+    cd azure-search-python-samples/Quickstart-Agentic-Retrieval
+    code .
     ```
 
-## Run the code
-
-To create and run the agentic retrieval pipeline:
+1. In `sample.env`, replace the placeholder values for `SEARCH_ENDPOINT` and `AOAI_ENDPOINT` with the URLs you obtained in [Get endpoints](#get-endpoints).
 
-1. Run the following command to install the required packages.
+1. Rename `sample.env` to `.env`.
 
-    ```console
-    pip install azure-identity requests azure-search-documents --pre
+    ```bash
+    mv sample.env .env
     ```
 
-1. Create a file named `agentic-retrieval.py` in the `quickstart-agentic-retrieval` folder.
+1. Open `quickstart-agentic-retrieval.ipynb`.
 
-1. Paste the following code into the file.
+1. Press **Ctrl+Shift+P**, select **Notebook: Select Notebook Kernel**, and follow the prompts to create a virtual environment. Select **requirements.txt** for the dependencies.
 
-    ```python
-    from azure.identity import DefaultAzureCredential, get_bearer_token_provider
-    from azure.search.documents.indexes.models import SearchIndex, SearchField, VectorSearch, VectorSearchProfile, HnswAlgorithmConfiguration, AzureOpenAIVectorizer, AzureOpenAIVectorizerParameters, SemanticSearch, SemanticConfiguration, SemanticPrioritizedFields, SemanticField, SearchIndexKnowledgeSource, SearchIndexKnowledgeSourceParameters, SearchIndexFieldReference, KnowledgeBase, KnowledgeBaseAzureOpenAIModel, KnowledgeSourceReference, KnowledgeRetrievalOutputMode, KnowledgeRetrievalLowReasoningEffort
-    from azure.search.documents.indexes import SearchIndexClient
-    from azure.search.documents import SearchIndexingBufferedSender
-    from azure.search.documents.knowledgebases import KnowledgeBaseRetrievalClient
-    from azure.search.documents.knowledgebases.models import KnowledgeBaseRetrievalRequest, KnowledgeBaseMessage, KnowledgeBaseMessageTextContent, SearchIndexKnowledgeSourceParams
-    import requests
-    import json
-    
-    # Define variables
-    search_endpoint = "PUT-YOUR-SEARCH-SERVICE-URL-HERE"
-    aoai_endpoint = "PUT-YOUR-AOAI-FOUNDRY-URL-HERE"
-    aoai_embedding_model = "text-embedding-3-large"
-    aoai_embedding_deployment = "text-embedding-3-large"
-    aoai_gpt_model = "gpt-5-mini"
-    aoai_gpt_deployment = "gpt-5-mini"
-    index_name = "earth-at-night"
-    knowledge_source_name = "earth-knowledge-source"
-    knowledge_base_name = "earth-knowledge-base"
-    search_api_version = "2025-11-01-preview"
-    
-    credential = DefaultAzureCredential()
-    token_provider = get_bearer_token_provider(credential, "https://search.azure.com/.default")
-    
-    # Create an index
-    azure_openai_token_provider = get_bearer_token_provider(credential, "https://cognitiveservices.azure.com/.default")
-
-    index = SearchIndex(
-        name=index_name,
-        fields=[
-            SearchField(name="id", type="Edm.String", key=True, filterable=True, sortable=True, facetable=True),
-            SearchField(name="page_chunk", type="Edm.String", filterable=False, sortable=False, facetable=False),
-            SearchField(name="page_embedding_text_3_large", type="Collection(Edm.Single)", stored=False, vector_search_dimensions=3072, vector_search_profile_name="hnsw_text_3_large"),
-            SearchField(name="page_number", type="Edm.Int32", filterable=True, sortable=True, facetable=True)
-        ],
-        vector_search=VectorSearch(
-            profiles=[VectorSearchProfile(name="hnsw_text_3_large", algorithm_configuration_name="alg", vectorizer_name="azure_openai_text_3_large")],
-            algorithms=[HnswAlgorithmConfiguration(name="alg")],
-            vectorizers=[
-                AzureOpenAIVectorizer(
-                    vectorizer_name="azure_openai_text_3_large",
-                    parameters=AzureOpenAIVectorizerParameters(
-                        resource_url=aoai_endpoint,
-                        deployment_name=aoai_embedding_deployment,
-                        model_name=aoai_embedding_model
-                    )
-                )
-            ]
-        ),
-        semantic_search=SemanticSearch(
-            default_configuration_name="semantic_config",
-            configurations=[
-                SemanticConfiguration(
-                    name="semantic_config",
-                    prioritized_fields=SemanticPrioritizedFields(
-                        content_fields=[
-                            SemanticField(field_name="page_chunk")
-                        ]
-                    )
-                )
-            ]
-        )
-    )
-    
-    index_client = SearchIndexClient(endpoint=search_endpoint, credential=credential)
-    index_client.create_or_update_index(index)
-    print(f"Index '{index_name}' created or updated successfully.")
-    
-    # Upload documents
-    url = "https://raw.githubusercontent.com/Azure-Samples/azure-search-sample-data/refs/heads/main/nasa-e-book/earth-at-night-json/documents.json"
-    documents = requests.get(url).json()
-    
-    with SearchIndexingBufferedSender(endpoint=search_endpoint, index_name=index_name, credential=credential) as client:
-        client.upload_documents(documents=documents)
-    
-    print(f"Documents uploaded to index '{index_name}' successfully.")
-    
-    # Create a knowledge source
-    ks = SearchIndexKnowledgeSource(
-        name=knowledge_source_name,
-        description="Knowledge source for Earth at night data",
-        search_index_parameters=SearchIndexKnowledgeSourceParameters(
-            search_index_name=index_name,
-            source_data_fields=[SearchIndexFieldReference(name="id"), SearchIndexFieldReference(name="page_number")]
-        ),
-    )
-    
-    index_client = SearchIndexClient(endpoint=search_endpoint, credential=credential)
-    index_client.create_or_update_knowledge_source(knowledge_source=ks)
-    print(f"Knowledge source '{knowledge_source_name}' created or updated successfully.")
-    
-    # Create a knowledge base
-    aoai_params = AzureOpenAIVectorizerParameters(
-        resource_url=aoai_endpoint,
-        deployment_name=aoai_gpt_deployment,
-        model_name=aoai_gpt_model,
-    )
-    
-    knowledge_base = KnowledgeBase(
-        name=knowledge_base_name,
-        models=[KnowledgeBaseAzureOpenAIModel(azure_open_ai_parameters=aoai_params)],
-        knowledge_sources=[
-            KnowledgeSourceReference(
-                name=knowledge_source_name
-            )
-        ],
-        output_mode=KnowledgeRetrievalOutputMode.ANSWER_SYNTHESIS,
-        answer_instructions="Provide a two sentence concise and informative answer based on the retrieved documents."
-    )
-    
-    index_client = SearchIndexClient(endpoint=search_endpoint, credential=credential)
-    index_client.create_or_update_knowledge_base(knowledge_base)
-    print(f"Knowledge base '{knowledge_base_name}' created or updated successfully.")
-    
-    # Set up messages
-    instructions = """
-    A Q&A agent that can answer questions about the Earth at night.
-    If you don't have the answer, respond with "I don't know".
-    """
-    
-    messages = [
-        {
-            "role": "system",
-            "content": instructions
-        }
-    ]
-    
-    # Run agentic retrieval
-    agent_client = KnowledgeBaseRetrievalClient(endpoint=search_endpoint, knowledge_base_name=knowledge_base_name, credential=credential)
-    query_1 = """
-        Why do suburban belts display larger December brightening than urban cores even though absolute light levels are higher downtown?
-        Why is the Phoenix nighttime street grid is so sharply visible from space, whereas large stretches of the interstate between midwestern cities remain comparatively dim?
-        """
-    
-    messages.append({
-        "role": "user",
-        "content": query_1
-    })
-    
-    req = KnowledgeBaseRetrievalRequest(
-        messages=[
-            KnowledgeBaseMessage(
-                role=m["role"],
-                content=[KnowledgeBaseMessageTextContent(text=m["content"])]
-            ) for m in messages if m["role"] != "system"
-        ],
-        knowledge_source_params=[
-            SearchIndexKnowledgeSourceParams(
-                knowledge_source_name=knowledge_source_name,
-                include_references=True,
-                include_reference_source_data=True,
-                always_query_source=True
-            )
-        ],
-        include_activity=True,
-        retrieval_reasoning_effort=KnowledgeRetrievalLowReasoningEffort
-    )
-    
-    result = agent_client.retrieve(retrieval_request=req)
-    print(f"Retrieved content from '{knowledge_base_name}' successfully.")
-    
-    # Display the response, activity, and references
-    response_contents = []
-    activity_contents = []
-    references_contents = []
-    
-    response_parts = []
-    for resp in result.response:
-        for content in resp.content:
-            response_parts.append(content.text)
-    response_content = "\n\n".join(response_parts) if response_parts else "No response found on 'result'"
-    
-    response_contents.append(response_content)
-    
-    # Print the three string values
-    print("response_content:\n", response_content, "\n")
+   When complete, you should see a `.venv` folder in the project directory.
 
-    messages.append({
-        "role": "assistant",
-        "content": response_content
-    })
-    
-    if result.activity:
-        activity_content = json.dumps([a.as_dict() for a in result.activity], indent=2)
-    else:
-        activity_content = "No activity found on 'result'"
-        
-    activity_contents.append(activity_content)
-    print("activity_content:\n", activity_content, "\n")
-    
-    if result.references:
-        references_content = json.dumps([r.as_dict() for r in result.references], indent=2)
-    else:
-        references_content = "No references found on 'result'"
-        
-    references_contents.append(references_content)
-    print("references_content:\n", references_content)
-    
-    # Continue the conversation
-    query_2 = "How do I find lava at night?"
-    messages.append({
-        "role": "user",
-        "content": query_2
-    })
-    
-    req = KnowledgeBaseRetrievalRequest(
-        messages=[
-            KnowledgeBaseMessage(
-                role=m["role"],
-                content=[KnowledgeBaseMessageTextContent(text=m["content"])]
-            ) for m in messages if m["role"] != "system"
-        ],
-        knowledge_source_params=[
-            SearchIndexKnowledgeSourceParams(
-                knowledge_source_name=knowledge_source_name,
-                include_references=True,
-                include_reference_source_data=True,
-                always_query_source=True
-            )
-        ],
-        include_activity=True,
-        retrieval_reasoning_effort=KnowledgeRetrievalLowReasoningEffort
-    )
-    
-    result = agent_client.retrieve(retrieval_request=req)
-    print(f"Retrieved content from '{knowledge_base_name}' successfully.")
-    
-    # Display the new retrieval response, activity, and references
-    response_parts = []
-    for resp in result.response:
-        for content in resp.content:
-            response_parts.append(content.text)
-    response_content = "\n\n".join(response_parts) if response_parts else "No response found on 'result'"
-    
-    response_contents.append(response_content)
-    
-    # Print the three string values
-    print("response_content:\n", response_content, "\n")
-    
-    if result.activity:
-        activity_content = json.dumps([a.as_dict() for a in result.activity], indent=2)
-    else:
-        activity_content = "No activity found on 'result'"
-        
-    activity_contents.append(activity_content)
-    print("activity_content:\n", activity_content, "\n")
-    
-    if result.references:
-        references_content = json.dumps([r.as_dict() for r in result.references], indent=2)
-    else:
-        references_content = "No references found on 'result'"
-        
-    references_contents.append(references_content)
-    print("references_content:\n", references_content)
-    
-    # Clean up resources
-    index_client = SearchIndexClient(endpoint=search_endpoint, credential=credential)
-    index_client.delete_knowledge_base(knowledge_base_name)
-    print(f"Knowledge base '{knowledge_base_name}' deleted successfully.")
-    
-    index_client = SearchIndexClient(endpoint=search_endpoint, credential=credential)
-    index_client.delete_knowledge_source(knowledge_source=knowledge_source_name)
-    print(f"Knowledge source '{knowledge_source_name}' deleted successfully.")
-    
-    index_client = SearchIndexClient(endpoint=search_endpoint, credential=credential)
-    index_client.delete_index(index_name)
-    print(f"Index '{index_name}' deleted successfully.")
+1. For keyless authentication with Microsoft Entra ID, sign in to your Azure account. If you have multiple subscriptions, select the one that contains your Azure AI Search and Microsoft Foundry resources.
+
+    ```azurecli
+    az login
     ```
 
-1. Set `search_endpoint` and `aoai_endpoint` to the values you obtained in [Get endpoints](#get-endpoints).
+## Run the code
 
-1. Run the following command to execute the code.
+1. Run the `Load connections` cell to install the required packages and load environment variables.
 
-    ```console
-    python agentic-retrieval.py
-    ```
+1. Run the remaining cells sequentially to create an index, upload documents, configure a knowledge source and knowledge base, and run agentic retrieval queries.
 
 ### Output
 
-The output of the code should look similar to the following:
+Each code cell prints its output to the notebook. The following example is the output after running all cells:
 
 ```
 Documents uploaded to index 'earth-at-night' successfully.
@@ -379,28 +116,7 @@ activity_content:
       "search": "December brightening in satellite nighttime lights: why do suburban belts show larger relative increases in December than urban cores despite higher absolute downtown light levels?"
     }
   },
-  {
-    "id": 2,
-    "type": "searchIndex",
-    "elapsed_ms": 632,
-    "knowledge_source_name": "earth-knowledge-source",
-    "query_time": "2025-11-05T16:17:48.985Z",
-    "count": 10,
-    "search_index_arguments": {
-      "search": "Why is Phoenix's nighttime street grid so sharply visible from space? factors: street-light layout, lamp type, urban form, light scattering, and satellite sensor characteristics in Phoenix, Arizona."
-    }
-  },
-  {
-    "id": 3,
-    "type": "searchIndex",
-    "elapsed_ms": 420,
-    "knowledge_source_name": "earth-knowledge-source",
-    "query_time": "2025-11-05T16:17:49.406Z",
-    "count": 11,
-    "search_index_arguments": {
-      "search": "Why are long stretches of interstate highways between Midwestern cities comparatively dim in satellite nighttime images? factors: highway lighting design, lamp spacing and type, vehicle headlights vs fixed lighting, and detection limits of nighttime sensors"
-    }
-  },
+  ... // Trimmed for brevity
   {
     "id": 4,
     "type": "agenticReasoning",
@@ -454,6 +170,8 @@ Index 'earth-at-night' deleted successfully.
 
 ## Understand the code
 
+[!INCLUDE [understand code note](../understand-code-note.md)]
+
 Now that you've run the code, let's break down the key steps:
 
 1. [Create a search index](#create-a-search-index)
@@ -582,7 +300,7 @@ knowledge_base = KnowledgeBase(
         )
     ],
     output_mode=KnowledgeRetrievalOutputMode.ANSWER_SYNTHESIS,
-    answer_instructions="Provide a two sentence concise and informative answer based on the retrieved documents."
+    answer_instructions="Provide a 2 sentence concise and informative answer based on the retrieved documents."
 )
 
 index_client = SearchIndexClient(endpoint=search_endpoint, credential=credential)
@@ -783,7 +501,7 @@ print("references_content:\n", references_content)
 
 [!INCLUDE [clean up resources (paid)](../resource-cleanup-paid.md)]
 
-Otherwise, the following code from `agentic-retrieval.py` deleted the objects you created in this quickstart.
+Otherwise, the following code from `quickstart-agentic-retrieval.ipynb` deleted the objects you created in this quickstart.
 
 ### Delete the knowledge base
 

@@ -4,42 +4,26 @@ author: haileytap
 ms.author: haileytapia
 ms.service: azure-ai-search
 ms.topic: include
-ms.date: 12/22/2025
+ms.date: 02/23/2026
 ---
 
 ## Configure access
 
-Before you begin, make sure you have permissions to access content and operations. We recommend Microsoft Entra ID for authentication and role-based access for authorization. You must be an **Owner** or **User Access Administrator** to assign roles. If roles aren't feasible, use [key-based authentication](../../search-security-api-keys.md) instead.
+Before you begin, make sure you have permissions to access content and operations. This quickstart uses Microsoft Entra ID for authentication and role-based access for authorization. You must be an **Owner** or **User Access Administrator** to assign roles. If roles aren't feasible, use [key-based authentication](../../search-security-api-keys.md) instead.
 
-To configure access for this quickstart, select both of the following tabs.
+To configure access for this quickstart:
 
-### [Azure AI Search](#tab/search)
+1. Sign in to the [Azure portal](https://portal.azure.com).
 
-Azure AI Search provides the agentic retrieval pipeline. Configure access for yourself and your search service to read and write data, interact with Foundry, and run the pipeline.
+1. On your Azure AI Search service:
 
-On your Azure AI Search service:
+    1. [Enable role-based access](../../search-security-enable-roles.md).
 
-1. [Enable role-based access](../../search-security-enable-roles.md).
+    1. [Create a system-assigned managed identity](../../search-how-to-managed-identities.md#create-a-system-managed-identity).
 
-1. [Create a system-assigned managed identity](../../search-how-to-managed-identities.md#create-a-system-managed-identity).
+    1. [Assign the following roles](../../search-security-rbac.md) to your user account: **Search Service Contributor**, **Search Index Data Contributor**, and **Search Index Data Reader**.
 
-1. [Assign the following roles](../../search-security-rbac.md) to yourself.
-
-    + **Search Service Contributor**
-
-    + **Search Index Data Contributor**
-
-    + **Search Index Data Reader**
-
-### [Microsoft Foundry](#tab/foundry)
-
-Microsoft Foundry provides the Azure OpenAI models used for embeddings, query planning, and answer generation. Grant your search service permission to use these models.
-
-On your Microsoft Foundry resource:
-
-+ Assign **Cognitive Services User** to the managed identity of your search service.
-
----
+1. On your Microsoft Foundry resource, assign **Cognitive Services User** to the managed identity of your search service.
 
 > [!IMPORTANT]
 > Agentic retrieval has two token-based billing models:
@@ -53,34 +37,18 @@ On your Microsoft Foundry resource:
 
 Each Azure AI Search service and Microsoft Foundry resource has an *endpoint*, which is a unique URL that identifies and provides network access to the resource. In a later section, you specify these endpoints to connect to your resources programmatically.
 
-To get the endpoints for this quickstart, select both of the following tabs.
-
-### [Azure AI Search](#tab/search)
+To get the endpoints for this quickstart:
 
-1. Sign in to the [Azure portal](https://portal.azure.com/) and select your search service.
-
-1. From the left pane, select **Overview**.
-
-1. Make a note of the endpoint, which should look like `https://my-service.search.windows.net`.
-
-### [Microsoft Foundry](#tab/foundry)
-
-1. Sign in to the [Azure portal](https://portal.azure.com/) and select your Microsoft Foundry resource.
-
-1. From the left pane, select **Resource Management** > **Keys and Endpoint**.
-
-1. Select the **OpenAI** tab.
-
-1. Make a note of the endpoint, which should look like `https://my-resource.openai.azure.com`.
-
----
+1. Sign in to the [Azure portal](https://portal.azure.com).
 
-## Deploy models
+1. On your Azure AI Search service:
 
-For this quickstart, you must deploy two Azure OpenAI models to your Microsoft Foundry project:
+    1. From the left pane, select **Overview**.
+    
+    1. Copy the URL, which should look like `https://my-service.search.windows.net`.
 
-+ An embedding model for text-to-vector conversion. This quickstart uses `text-embedding-3-large`, but you can use any `text-embedding` model.
+1. On your Microsoft Foundry resource:
 
-+ An LLM for query planning and answer generation. This quickstart uses `gpt-5-mini`, but you can use any [supported LLM for agentic retrieval](../../agentic-retrieval-how-to-create-knowledge-base.md#supported-models).
+    1. From the left pane, select **Resource Management** > **Keys and Endpoint**.
 
-For deployment instructions, see [Deploy Microsoft Foundry Models in the Foundry portal](/azure/ai-foundry/how-to/deploy-models-openai).
+    1. Copy the URL on the **OpenAI** tab, which should look like `https://my-resource.openai.azure.com/`.

@@ -4,7 +4,7 @@ author: haileytap
 ms.author: haileytapia
 ms.service: azure-ai-search
 ms.topic: include
-ms.date: 01/14/2026
+ms.date: 02/23/2026
 ms.custom: dev-focus
 ai-usage: ai-assisted
 ---
@@ -15,7 +15,10 @@ In this quickstart, you use [agentic retrieval](../../agentic-retrieval-overview
 
 A *knowledge base* orchestrates agentic retrieval by decomposing complex queries into subqueries, running the subqueries against one or more *knowledge sources*, and returning results with metadata. By default, the knowledge base outputs raw content from your sources, but this quickstart uses the answer synthesis output mode for natural-language answer generation.
 
-Although you can provide your own data, this quickstart uses [sample JSON documents](https://github.com/Azure-Samples/azure-search-sample-data/tree/main/nasa-e-book/earth-at-night-json) from NASA's Earth at Night e-book. The documents describe general science topics and images of Earth at night as observed from space.
+Although you can use your own data, this quickstart uses [sample JSON documents](https://github.com/Azure-Samples/azure-search-sample-data/tree/main/nasa-e-book/earth-at-night-json) from NASA's Earth at Night e-book.
+
+> [!TIP]
+> Source code for the TypeScript version of this quickstart isn't available. You can copy code directly from this article.
 
 ## Prerequisites
 
@@ -25,68 +28,66 @@ Although you can provide your own data, this quickstart uses [sample JSON docume
 
 + A [Microsoft Foundry project](/azure/ai-foundry/how-to/create-projects) and resource. When you create a project, the resource is automatically created.
 
-+ The [Azure CLI](/cli/azure/install-azure-cli) for keyless authentication with Microsoft Entra ID.
++ An embedding model [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai) for text-to-vector conversion. You can use any `text-embedding` model, such as `text-embedding-3-large`.
 
-+ [Visual Studio Code](https://code.visualstudio.com/download) and the latest LTS version of [Node.js](https://nodejs.org/en/download/).
++ An LLM [deployed to your project](/azure/ai-foundry/how-to/deploy-models-openai) for query planning and answer generation. You can use any [supported LLM](../../agentic-retrieval-how-to-create-knowledge-base.md#supported-models), such as `gpt-5-mini`.
 
-[!INCLUDE [Setup](./agentic-retrieval-setup.md)]
++ [Node.js 20 LTS](https://nodejs.org/en/download/) or later.
 
-## Set up the environment
++ [Visual Studio Code](https://code.visualstudio.com/download).
 
-To set up the console application for this quickstart:
++ The [Azure CLI](/cli/azure/install-azure-cli) for keyless authentication with Microsoft Entra ID.
+
+[!INCLUDE [agentic retrieval setup](agentic-retrieval-setup.md)]
+
+## Set up the environment
 
 1. Create a folder named `quickstart-agentic-retrieval` to contain the application.
 
 1. Open the folder in Visual Studio Code.
 
-1. Select **Terminal** > **New Terminal**, and then run the following commands to initialize the `package.json` file.
+1. Select **Terminal** > **New Terminal**, and then run the following commands to initialize `package.json`.
 
-    ```console
+    ```bash
     npm init -y
     npm pkg set type=module
     ```
 
 1. Install TypeScript as a development dependency.
 
-    ```console
+    ```bash
     npm install --save-dev typescript @types/node
     ```
 
 1. Install the [Azure AI Search client library for JavaScript](/javascript/api/overview/azure/search-documents-readme).
 
-    ```console
+    ```bash
     npm install @azure/search-documents@12.3.0-beta.1
     ```
 
 1. For keyless authentication with Microsoft Entra ID, install the [Azure Identity client library for JavaScript](/javascript/api/overview/azure/identity-readme).
 
-    ```console
+    ```bash
     npm install @azure/identity
     ```
 
-1. For keyless authentication with Microsoft Entra ID, sign in to your Azure account. If you have multiple subscriptions, select the one that contains your Azure AI Search service and Microsoft Foundry project.
+1. For keyless authentication with Microsoft Entra ID, sign in to your Azure account. If you have multiple subscriptions, select the one that contains your Azure AI Search and Microsoft Foundry resources.
 
-    ```console
+    ```bash
     az login
     ```
 
 ## Run the code
 
-To create and run the agentic retrieval pipeline:
-
-1. Create a file named `.env` in the `quickstart-agentic-retrieval` folder.
-
-1. Paste the following environment variables into the `.env` file.
+1. Create a file named `.env` in the `quickstart-agentic-retrieval` folder, and then paste the following content. Replace the placeholder values with the URLs you obtained in [Get endpoints](#get-endpoints).
 
     ```
     AZURE_SEARCH_ENDPOINT = https://<your-search-service-name>.search.windows.net
-    AZURE_OPENAI_ENDPOINT = https://<your-ai-foundry-resource-name>.openai.azure.com/
+    AZURE_OPENAI_ENDPOINT = https://<your-foundry-resource-name>.openai.azure.com/
     AZURE_OPENAI_GPT_DEPLOYMENT = gpt-5-mini
     AZURE_OPENAI_EMBEDDING_DEPLOYMENT = text-embedding-3-large
     ```
 
-1. Set `AZURE_SEARCH_ENDPOINT` and `AZURE_OPENAI_ENDPOINT` to the values you obtained in [Get endpoints](#get-endpoints).
-
 1. Create a file named `index.ts`, and then paste the following code into the file.
 
     ```typescript
@@ -454,21 +455,21 @@ To create and run the agentic retrieval pipeline:
 
 1. Transpile the code from TypeScript to JavaScript.
 
-    ```console
+    ```bash
     npx tsc
     ```
 
 1. Build and run the application.
 
-    ```console
+    ```bash
     node --env-file ./.env dist/index.js
     ```
 
 ### Output
 
 The output of the application should be similar to the following:
 
-```console
+```output
 Waiting for indexing to complete...
 Expected documents: 194
 Current indexed count: 194
@@ -516,58 +517,7 @@ Activity Type: searchIndex
     "semanticConfigurationName": "semantic_config"
   }
 }
-Activity Type: searchIndex
-{
-  "id": 2,
-  "type": "searchIndex",
-  "elapsedMs": 538,
-  "knowledgeSourceName": "earth-knowledge-source",
-  "queryTime": "2025-12-19T15:38:24.001Z",
-  "count": 0,
-  "searchIndexArguments": {
-    "search": "factors that make Phoenix nighttime street grid highly visible from space reasons highway/interstate lighting visibility differences Midwestern interstates dim",
-    "filter": null,
-    "sourceDataFields": [
-      {
-        "name": "page_chunk"
-      },
-      {
-        "name": "id"
-      },
-      {
-        "name": "page_number"
-      }
-    ],
-    "searchFields": [],
-    "semanticConfigurationName": "semantic_config"
-  }
-}
-Activity Type: searchIndex
-{
-  "id": 3,
-  "type": "searchIndex",
-  "elapsedMs": 465,
-  "knowledgeSourceName": "earth-knowledge-source",
-  "queryTime": "2025-12-19T15:38:24.467Z",
-  "count": 2,
-  "searchIndexArguments": {
-    "search": "satellite nighttime lights seasonal variations suburban brightening studies December holiday lighting residential vs commercial lighting patterns",
-    "filter": null,
-    "sourceDataFields": [
-      {
-        "name": "page_chunk"
-      },
-      {
-        "name": "id"
-      },
-      {
-        "name": "page_number"
-      }
-    ],
-    "searchFields": [],
-    "semanticConfigurationName": "semantic_config"
-  }
-}
+... // Trimmed for brevity
 Activity Type: agenticReasoning
 {
   "id": 4,
@@ -773,7 +723,7 @@ console.log(`✓ All ${documents.length} documents indexed successfully!`);
 
 A knowledge source is a reusable reference to source data. The following code defines a knowledge source named `earth-knowledge-source` that targets the `earth-at-night` index.
 
-`source_data_fields` specifies which index fields are included in citation references. This example includes only human-readable fields to avoid lengthy, uninterpretable embeddings in responses.
+`sourceDataFields` specifies which index fields are included in citation references. This example includes only human-readable fields to avoid lengthy, uninterpretable embeddings in responses.
 
 ```typescript
 await searchIndexClient.createKnowledgeSource({

@@ -71,17 +71,29 @@ Full-text search uses Apache Lucene for indexing and queries and the BM25 rankin
 
 Compile and run the application.
 
+### [Windows](#tab/windows)
+
 ```bash
 javac -d target/classes -cp "target/dependency/*" src/main/java/azure/search/sample/*.java
 java -cp "target/classes;target/dependency/*" azure.search.sample.App
 ```
 
-On macOS or Linux, use `:` instead of `;` as the classpath separator.
+### [macOS](#tab/macos)
+
+```bash
+javac -d target/classes -cp "target/dependency/*" src/main/java/azure/search/sample/*.java
+java -cp "target/classes:target/dependency/*" azure.search.sample.App
+```
+
+### [Linux](#tab/linux)
 
 ```bash
+javac -d target/classes -cp "target/dependency/*" src/main/java/azure/search/sample/*.java
 java -cp "target/classes:target/dependency/*" azure.search.sample.App
 ```
 
+---
+
 ### Output
 
 The output should be similar to the following:

@@ -78,7 +78,7 @@ In Azure AI Search, a vector index has an index schema that defines vector and n
 
 ## Run the code
 
-1. Run the `Install packages and set variables` cell to load environment variables and verify the package installation.
+1. Run the `Install packages and set variables` cell to install the required packages and load environment variables.
 
 1. Run the remaining cells sequentially to create a vector index, upload documents, and run different types of vector queries.
 

@@ -12,7 +12,7 @@ ms.custom:
   - devx-track-extended-java
   - ignite-2023
 ms.topic: concept-article
-ms.date: 01/13/2026
+ms.date: 02/23/2026
 ---
 
 # Java samples for Azure AI Search
@@ -48,6 +48,7 @@ Code samples from the Azure AI Search team demonstrate features and workflows. T
 
 | Sample | Article | Description |
 |--|--|--|
+| [quickstart-agentic-retrieval](https://github.com/Azure-Samples/azure-search-java-samples/tree/main/quickstart-agentic-retrieval) | [Quickstart: Agentic retrieval](search-get-started-agentic-retrieval.md) | Integrate semantic ranking with LLM-powered query planning and answer generation. |
 | [quickstart-keyword-search](https://github.com/Azure-Samples/azure-search-java-samples/tree/main/quickstart-keyword-search) | [Quickstart: Full-text search](search-get-started-text.md) | Create, load, and query a search index using sample data. |
 | [quickstart-semantic-ranking](https://github.com/Azure-Samples/azure-search-java-samples/tree/main/quickstart-semantic-ranking) | [Quickstart: Semantic ranking](search-get-started-semantic.md) | Add semantic ranking to an index schema and run semantic queries. |
 | [quickstart-vector-search](https://github.com/Azure-Samples/azure-search-java-samples/tree/main/quickstart-vector-search) | [Quickstart: Vector search](search-get-started-vector.md) | Index and query vector content. |

@@ -6,7 +6,7 @@ author: haileytap
 ms.author: haileytapia
 ms.service: azure-ai-search
 ms.topic: quickstart
-ms.date: 01/14/2026
+ms.date: 02/23/2026
 ms.custom: dev-focus
 ai-usage: ai-assisted
 zone_pivot_groups: search-get-started-agentic-retrieval

@@ -7,7 +7,7 @@ ms.author: gimondra
 manager: nitinme
 ms.service: azure-ai-search
 ms.topic: how-to
-ms.date: 11/21/2025
+ms.date: 02/17/2026
 ms.custom:
   - build-2024
   - ignite-2024
@@ -44,7 +44,9 @@ This article uses the REST APIs to illustrate each step.
 
 + An AI Search service, basic pricing tier or higher, configured for either a [system managed identity](search-how-to-managed-identities.md#create-a-system-managed-identity) or [user-assigned assigned managed identity](search-how-to-managed-identities.md#create-a-user-assigned-managed-identity). The AI Search service must reside within the same tenant as the Microsoft Fabric workspace.
   
-+ A Contributor role assignment in the Microsoft Fabric workspace where the lakehouse is located. Steps are outlined in the [Grant permissions](#assign-service-permissions) section of this article.
++ An Administrator or Contributor role assignment in the Microsoft Fabric workspace where the lakehouse is located. Steps are outlined in the [Grant permissions](#assign-service-permissions) section of this article.
+
++ Allow [access to OneLake data from applications that are outside of the Fabric environment](/fabric/onelake/security/get-started-security#allow-apps-running-outside-of-fabric-to-access-data-via-onelake). 
 
 + A [REST client](search-get-started-text.md) to formulate REST calls similar to the ones shown in this article.
 
@@ -96,6 +98,10 @@ The following OneLake shortcuts are supported by the OneLake files indexer:
 
 + [Google Cloud Storage shortcut](/fabric/onelake/create-gcs-shortcut)
 
+
+> [!IMPORTANT]
+> Failing to meet any of the prerequisites, or attempting an operation covered by the documented limitations, will cause errors when listing items in the lakehouse.
+
 ## Prepare data for indexing
 
 Before you set up indexing, review your source data to determine whether any changes should be made to your data in the lakehouse. An indexer can index content from one container at a time. By default, all files in the container are processed. You have several options for more selective processing:

@@ -65,7 +65,7 @@ The following JSON document is composed of simple fields and complex fields. Com
 
 As with any index definition, you can use the Azure portal, [REST API](/rest/api/searchservice/indexes/create), or [.NET SDK](/dotnet/api/azure.search.documents.indexes.models.searchindex) to create a schema that includes complex types. 
 
-Other Azure SDKs provide samples in [Python](https://github.com/Azure/azure-sdk-for-python/blob/main/sdk/search/azure-search-documents/samples/sample_index_crud_operations.py), [Java](https://github.com/Azure/azure-sdk-for-java/blob/main/sdk/search/azure-search-documents/src/samples/java/com/azure/search/documents/indexes/CreateIndexExample.java), and [JavaScript](https://github.com/Azure/azure-sdk-for-js/blob/main/sdk/search/search-documents/samples/v11/javascript/indexOperations.js).
+Other Azure SDKs provide samples in [Python](https://github.com/Azure/azure-sdk-for-python/blob/main/sdk/search/azure-search-documents/samples/sample_index_crud.py), [Java](https://github.com/Azure/azure-sdk-for-java/blob/main/sdk/search/azure-search-documents/src/samples/java/com/azure/search/documents/indexes/CreateIndexExample.java), and [JavaScript](https://github.com/Azure/azure-sdk-for-js/blob/main/sdk/search/search-documents/samples/v11/javascript/indexOperations.js).
 
 ### [**Azure portal**](#tab/portal)