.Net: Add Bedrock Agent to .Net SDK

dotnet/Directory.Packages.props

@@ -5,6 +5,8 @@
     <ManagePackageVersionsCentrally>true</ManagePackageVersionsCentrally>
   </PropertyGroup>
   <ItemGroup>
+    <PackageVersion Include="AWSSDK.BedrockAgent" Version="3.7.416.4" />
+    <PackageVersion Include="AWSSDK.BedrockAgentRuntime" Version="3.7.418.3" />
     <PackageVersion Include="AWSSDK.BedrockRuntime" Version="4.0.0-preview.5" />
     <PackageVersion Include="AWSSDK.Core" Version="4.0.0-preview.5" />
     <PackageVersion Include="AWSSDK.Extensions.NETCore.Setup" Version="3.7.301" />
@@ -60,6 +62,7 @@
     <PackageVersion Include="System.IdentityModel.Tokens.Jwt" Version="7.5.1" />
     <PackageVersion Include="System.IO.Packaging" Version="8.0.1" />
     <PackageVersion Include="System.Linq.Async" Version="6.0.1" />
+    <PackageVersion Include="System.Memory.Data" Version="8.0.1" />
     <PackageVersion Include="System.Net.Http" Version="4.3.4" />
     <PackageVersion Include="System.Numerics.Tensors" Version="8.0.0" />
     <PackageVersion Include="System.Text.Json" Version="8.0.5" />

dotnet/SK-dotnet.sln

@@ -445,6 +445,8 @@ Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "Plugins.AI.UnitTests", "src
 EndProject
 Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "Connectors.Postgres.UnitTests", "src\Connectors\Connectors.Postgres.UnitTests\Connectors.Postgres.UnitTests.csproj", "{2A1EC0DA-AD01-4421-AADC-1DFF65C71CCC}"
 EndProject
+Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "Agents.Bedrock", "src\Agents\Bedrock\Agents.Bedrock.csproj", "{8C658E1E-83C8-4127-B8BF-27A638A45DDD}"
+EndProject
 Global
 	GlobalSection(SolutionConfigurationPlatforms) = preSolution
 		Debug|Any CPU = Debug|Any CPU
@@ -1196,6 +1198,12 @@ Global
 		{2A1EC0DA-AD01-4421-AADC-1DFF65C71CCC}.Publish|Any CPU.Build.0 = Debug|Any CPU
 		{2A1EC0DA-AD01-4421-AADC-1DFF65C71CCC}.Release|Any CPU.ActiveCfg = Release|Any CPU
 		{2A1EC0DA-AD01-4421-AADC-1DFF65C71CCC}.Release|Any CPU.Build.0 = Release|Any CPU
+		{8C658E1E-83C8-4127-B8BF-27A638A45DDD}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
+		{8C658E1E-83C8-4127-B8BF-27A638A45DDD}.Debug|Any CPU.Build.0 = Debug|Any CPU
+		{8C658E1E-83C8-4127-B8BF-27A638A45DDD}.Publish|Any CPU.ActiveCfg = Publish|Any CPU
+		{8C658E1E-83C8-4127-B8BF-27A638A45DDD}.Publish|Any CPU.Build.0 = Publish|Any CPU
+		{8C658E1E-83C8-4127-B8BF-27A638A45DDD}.Release|Any CPU.ActiveCfg = Release|Any CPU
+		{8C658E1E-83C8-4127-B8BF-27A638A45DDD}.Release|Any CPU.Build.0 = Release|Any CPU
 	EndGlobalSection
 	GlobalSection(SolutionProperties) = preSolution
 		HideSolutionNode = FALSE
@@ -1360,6 +1368,7 @@ Global
 		{0C64EC81-8116-4388-87AD-BA14D4B59974} = {D6D598DF-C17C-46F4-B2B9-CDE82E2DE132}
 		{03ACF9DD-00C9-4F2B-80F1-537E2151AF5F} = {D6D598DF-C17C-46F4-B2B9-CDE82E2DE132}
 		{2A1EC0DA-AD01-4421-AADC-1DFF65C71CCC} = {5A7028A7-4DDF-4E4F-84A9-37CE8F8D7E89}
+		{8C658E1E-83C8-4127-B8BF-27A638A45DDD} = {6823CD5E-2ABE-41EB-B865-F86EC13F0CF9}
 	EndGlobalSection
 	GlobalSection(ExtensibilityGlobals) = postSolution
 		SolutionGuid = {FBDC56A3-86AD-4323-AA0F-201E59123B83}

dotnet/samples/Concepts/Concepts.csproj

@@ -58,6 +58,7 @@
     <ProjectReference Include="..\..\src\Agents\AzureAI\Agents.AzureAI.csproj" />
     <ProjectReference Include="..\..\src\Agents\Core\Agents.Core.csproj" />
     <ProjectReference Include="..\..\src\Agents\OpenAI\Agents.OpenAI.csproj" />
+    <ProjectReference Include="..\..\src\Agents\Bedrock\Agents.Bedrock.csproj" />
     <ProjectReference Include="..\..\src\Connectors\Connectors.Google\Connectors.Google.csproj" />
     <ProjectReference Include="..\..\src\Connectors\Connectors.HuggingFace\Connectors.HuggingFace.csproj" />
     <ProjectReference Include="..\..\src\Connectors\Connectors.Memory.AzureAISearch\Connectors.Memory.AzureAISearch.csproj" />

dotnet/samples/GettingStartedWithAgents/AzureAIAgent/Step05_AzureAIAgent_FileSearch.cs

@@ -9,7 +9,7 @@
 namespace GettingStarted.AzureAgents;
 
 /// <summary>
-/// Demonstrate using code-interpreter on <see cref="AzureAIAgent"/> .
+/// Demonstrate using <see cref="AzureAIAgent"/> with file search.
 /// </summary>
 public class Step05_AzureAIAgent_FileSearch(ITestOutputHelper output) : BaseAzureAgentTest(output)
 {

dotnet/samples/GettingStartedWithAgents/BedrockAgent/README.md

@@ -0,0 +1,38 @@
+# Concept samples on how to use AWS Bedrock agents
+
+## Pre-requisites
+
+1. You need to have an AWS account and [access to the foundation models](https://docs.aws.amazon.com/bedrock/latest/userguide/model-access-permissions.html)
+2. [AWS CLI installed](https://docs.aws.amazon.com/cli/latest/userguide/getting-started-install.html) and [configured](https://boto3.amazonaws.com/v1/documentation/api/latest/guide/quickstart.html#configuration)
+
+## Before running the samples
+
+You need to set up some user secrets to run the samples.
+
+### `BedrockAgent:AgentResourceRoleArn`
+
+On your AWS console, go to the IAM service and go to **Roles**. Find the role you want to use and click on it. You will find the ARN in the summary section.
+
+```
+dotnet user-secrets set "BedrockAgent:AgentResourceRoleArn" "arn:aws:iam::...:role/..."
+```
+
+### `BedrockAgent:FoundationModel`
+
+You need to make sure you have permission to access the foundation model. You can find the model ID in the [AWS documentation](https://docs.aws.amazon.com/bedrock/latest/userguide/models-supported.html). To see the models you have access to, find the policy attached to your role you should see a list of models you have access to under the `Resource` section.
+
+```
+dotnet user-secrets set "BedrockAgent:FoundationModel" "..."
+```
+
+### How to add the `bedrock:InvokeModelWithResponseStream` action to an IAM policy
+
+1. Open the [IAM console](https://console.aws.amazon.com/iam/).
+2. On the left navigation pane, choose `Roles` under `Access management`.
+3. Find the role you want to edit and click on it.
+4. Under the `Permissions policies` tab, click on the policy you want to edit.
+5. Under the `Permissions defined in this policy` section, click on the service. You should see **Bedrock** if you already have access to the Bedrock agent service.
+6. Click on the service, and then click `Edit`.
+7. On the right, you will be able to add an action. Find the service and search for `InvokeModelWithResponseStream`.
+8. Check the box next to the action and then scroll all the way down and click `Next`.
+9. Follow the prompts to save the changes.

dotnet/samples/GettingStartedWithAgents/BedrockAgent/Step01_BedrockAgent.cs

@@ -0,0 +1,73 @@
+// Copyright (c) Microsoft. All rights reserved.
+
+using Microsoft.SemanticKernel.Agents.Bedrock;
+using Microsoft.SemanticKernel.Agents.Bedrock.Extensions;
+
+namespace GettingStarted.BedrockAgents;
+
+/// <summary>
+/// This example demonstrates how to interact with a <see cref="BedrockAgent"/> in the most basic way.
+/// </summary>
+public class Step01_BedrockAgent(ITestOutputHelper output) : BaseBedrockAgentTest(output)
+{
+    private const string UserQuery = "Why is the sky blue in one sentence?";
+
+    /// <summary>
+    /// Demonstrates how to create a new <see cref="BedrockAgent"/> and interact with it.
+    /// The agent will respond to the user query.
+    /// </summary>
+    [Fact]
+    public async Task UseNewAgentAsync()
+    {
+        // Create the agent
+        var bedrockAgent = await this.CreateAgentAsync("Step01_BedrockAgent");
+
+        // Respond to user input
+        try
+        {
+            var responses = bedrockAgent.InvokeAsync(BedrockAgent.CreateSessionId(), UserQuery, null);
+            await foreach (var response in responses)
+            {
+                this.Output.WriteLine(response.Content);
+            }
+        }
+        finally
+        {
+            await this.Client.DeleteAgentAsync(new() { AgentId = bedrockAgent.Id });
+        }
+    }
+
+    /// <summary>
+    /// Demonstrates how to create a new <see cref="BedrockAgent"/> and interact with it using streaming.
+    /// The agent will respond to the user query.
+    /// </summary>
+    [Fact]
+    public async Task UseNewAgentStreamingAsync()
+    {
+        // Create the agent
+        var bedrockAgent = await this.CreateAgentAsync("Step01_BedrockAgent_Streaming");
+
+        // Respond to user input
+        try
+        {
+            var streamingResponses = bedrockAgent.InvokeStreamingAsync(BedrockAgent.CreateSessionId(), UserQuery, null);
+            await foreach (var response in streamingResponses)
+            {
+                this.Output.WriteLine(response.Content);
+            }
+        }
+        finally
+        {
+            await this.Client.DeleteAgentAsync(new() { AgentId = bedrockAgent.Id });
+        }
+    }
+
+    protected override async Task<BedrockAgent> CreateAgentAsync(string agentName)
+    {
+        // Create a new agent on the Bedrock Agent service and prepare it for use
+        var agentModel = await this.Client.CreateAndPrepareAgentAsync(this.GetCreateAgentRequest(agentName));
+        // Create a new BedrockAgent instance with the agent model and the client
+        // so that we can interact with the agent using Semantic Kernel contents.
+        return new BedrockAgent(agentModel, this.Client);
+    }
+}

dotnet/samples/GettingStartedWithAgents/BedrockAgent/Step02_BedrockAgent_CodeInterpreter.cs

@@ -0,0 +1,90 @@
+// Copyright (c) Microsoft. All rights reserved.
+
+using System.Reflection;
+using Microsoft.SemanticKernel;
+using Microsoft.SemanticKernel.Agents.Bedrock;
+using Microsoft.SemanticKernel.Agents.Bedrock.Extensions;
+
+namespace GettingStarted.BedrockAgents;
+
+/// <summary>
+/// This example demonstrates how to interact with a <see cref="BedrockAgent"/> with code interpreter enabled.
+/// </summary>
+public class Step02_BedrockAgent_CodeInterpreter(ITestOutputHelper output) : BaseBedrockAgentTest(output)
+{
+    private const string UserQuery = @"Create a bar chart for the following data:
+Panda   5
+Tiger   8
+Lion    3
+Monkey  6
+Dolphin  2";
+
+    /// <summary>
+    /// Demonstrates how to create a new <see cref="BedrockAgent"/> with code interpreter enabled and interact with it.
+    /// The agent will respond to the user query by creating a Python code that will be executed by the code interpreter.
+    /// The output of the code interpreter will be a file containing the bar chart, which will be returned to the user.
+    /// </summary>
+    [Fact]
+    public async Task UseAgentWithCodeInterpreterAsync()
+    {
+        // Create the agent
+        var bedrockAgent = await this.CreateAgentAsync("Step02_BedrockAgent_CodeInterpreter");
+
+        // Respond to user input
+        try
+        {
+            BinaryContent? binaryContent = null;
+            var responses = bedrockAgent.InvokeAsync(BedrockAgent.CreateSessionId(), UserQuery, null);
+            await foreach (var response in responses)
+            {
+                if (response.Content != null)
+                {
+                    this.Output.WriteLine(response.Content);
+                }
+                if (binaryContent == null && response.Items.Count > 0)
+                {
+                    binaryContent = response.Items.OfType<BinaryContent>().FirstOrDefault();
+                }
+            }
+
+            if (binaryContent == null)
+            {
+                throw new InvalidOperationException("No file found in the response.");
+            }
+
+            // Save the file to the same directory as the test assembly
+            var filePath = Path.Combine(
+                Path.GetDirectoryName(Assembly.GetExecutingAssembly().Location)!,
+                binaryContent.Metadata!["Name"]!.ToString()!);
+            this.Output.WriteLine($"Saving file to {filePath}");
+            binaryContent.WriteToFile(filePath, overwrite: true);
+
+            // Expected output:
+            // Here is the bar chart for the given data:
+            // [A bar chart showing the following data:
+            // Panda   5
+            // Tiger   8
+            // Lion    3
+            // Monkey  6
+            // Dolphin 2]
+            // Saving file to ...
+        }
+        finally
+        {
+            await this.Client.DeleteAgentAsync(new() { AgentId = bedrockAgent.Id });
+        }
+    }
+
+    protected override async Task<BedrockAgent> CreateAgentAsync(string agentName)
+    {
+        // Create a new agent on the Bedrock Agent service and prepare it for use
+        var agentModel = await this.Client.CreateAndPrepareAgentAsync(this.GetCreateAgentRequest(agentName));
+        // Create a new BedrockAgent instance with the agent model and the client
+        // so that we can interact with the agent using Semantic Kernel contents.
+        var bedrockAgent = new BedrockAgent(agentModel, this.Client);
+        // Create the code interpreter action group and prepare the agent for interaction
+        await bedrockAgent.CreateCodeInterpreterActionGroupAsync();
+
+        return bedrockAgent;
+    }
+}