Stop failing held issue triage runs

* feat(python): Add MCP client OTel spans per GenAI semantic conventions

Implement MCP client spans per the OTel GenAI Semantic Conventions for MCP
(https://opentelemetry.io/docs/specs/semconv/gen-ai/mcp/#client).

Operations instrumented:
- initialize: CLIENT span capturing MCP session setup
- tools/list: CLIENT span for tool listing (per-page)
- prompts/list: CLIENT span for prompt listing (per-page)
- tools/call: CLIENT span (nested under execute_tool when called via FunctionTool)
- prompts/get: CLIENT span

Span attributes follow the MCP semantic conventions:
- Required: mcp.method.name
- Conditional: error.type, gen_ai.tool.name, gen_ai.prompt.name
- Recommended: gen_ai.operation.name, mcp.protocol.version, mcp.session.id,
  network.transport, server.address, server.port

Transport-specific attributes per subclass:
- MCPStdioTool: network.transport=pipe
- MCPStreamableHTTPTool: network.transport=tcp, network.protocol.name=http
- MCPWebsocketTool: network.transport=tcp, network.protocol.name=websocket

All span creation gated behind OBSERVABILITY_SETTINGS.ENABLED.

Closes #3624
Closes #4697

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* refactor: simplify MCP spans — remove enrichment logic and protocol version caching

- Always create nested CLIENT spans for tools/call instead of enriching
  the parent execute_tool span
- Remove _ACTIVE_TOOL_EXECUTION_SPAN contextvar (no longer needed)
- Remove enrich_span_with_mcp_attributes() helper
- Remove _otel_error_type preservation in FunctionTool.invoke()
- Remove _mcp_protocol_version instance variable; protocol version is
  only set on the initialize span where it is available

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Refine copilot solution

* fix: enable automatic exception recording on MCP spans

Remove record_exception=False and set_status_on_exception=False from
create_mcp_client_span. Let OTel handle exception recording and status
setting automatically. The manual set_mcp_span_error calls for tools/call
still correctly set error.type (which OTel's automatic handling doesn't
touch), so tool_error is preserved.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Reduce number of lines

* Add comment to sample

* test: address PR review comments on MCP observability tests

- Fix initialize test to call mocked session.initialize() and read
  protocolVersion from the result instead of hardcoding it
- Add tools/call McpError error-path test
- Add prompts/get McpError error-path test

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Fix export error

---------

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

.github/workflows/issue-triage.yml

@@ -153,8 +153,7 @@ jobs:
         if: ${{ steps.spam.outputs.allow_triage != 'true' }}
         shell: bash
         run: |
-          echo "Stopping: issue triage preflight did not allow automation."
-          exit 1
+          echo "Issue triage preflight withheld automation; maintainer review is required."
 
       - name: Reproduce reported issue
         if: ${{ steps.spam.outputs.allow_triage == 'true' }}

dotnet/src/Microsoft.Agents.AI.Harness/HarnessAgent.cs

@@ -138,7 +138,7 @@ public sealed class HarnessAgent : DelegatingAIAgent
 
         if (options?.DisableToolApproval is not true)
         {
-            builder.UseToolApproval();
+            builder.UseToolApproval(options?.ToolApprovalAgentOptions);
         }
 
         if (options?.DisableOpenTelemetry is not true)

dotnet/src/Microsoft.Agents.AI.Harness/HarnessAgentOptions.cs

@@ -101,6 +101,15 @@ public sealed class HarnessAgentOptions
     /// </remarks>
     public bool DisableToolApproval { get; set; }
 
+    /// <summary>
+    /// Gets or sets the options for the <see cref="ToolApprovalAgent"/> middleware.
+    /// </summary>
+    /// <remarks>
+    /// When <see langword="null"/>, the <see cref="ToolApprovalAgent"/> uses default settings.
+    /// This property has no effect when <see cref="DisableToolApproval"/> is <see langword="true"/>.
+    /// </remarks>
+    public ToolApprovalAgentOptions? ToolApprovalAgentOptions { get; set; }
+
     /// <summary>
     /// Gets or sets a value indicating whether the <see cref="FileMemoryProvider"/> is disabled.
     /// </summary>

dotnet/src/Microsoft.Agents.AI.Workflows.Declarative/ObjectModel/ForeachExecutor.cs

@@ -49,7 +49,7 @@ internal sealed class ForeachExecutor : DeclarativeActionExecutor<Foreach>
         EvaluationResult<DataValue> expressionResult = this.Evaluator.GetValue(this.Model.Items);
         if (expressionResult.Value is TableDataValue tableValue)
         {
-            this._values = [.. tableValue.Values.Select(value => value.Properties.Values.First().ToFormula())];
+            this._values = [.. tableValue.Values.Select(value => value.ToFormula())];
         }
         else
         {

dotnet/src/Microsoft.Agents.AI/ChatClient/ChatClientAgentOptions.cs

@@ -181,6 +181,36 @@ public sealed class ChatClientAgentOptions
     [Experimental(DiagnosticIds.Experiments.AgentsAIExperiments)]
     public bool EnableMessageInjection { get; set; }
 
+    /// <summary>
+    /// Gets or sets a value indicating whether to store automatically approved function calls in the session state
+    /// for tools that do not require approval when they are returned alongside tools that do.
+    /// </summary>
+    /// <remarks>
+    /// <para>
+    /// <see cref="FunctionInvokingChatClient"/> has an all-or-nothing behavior for approvals: when any tool
+    /// in a response is an <see cref="ApprovalRequiredAIFunction"/>, it converts all <see cref="FunctionCallContent"/>
+    /// items to <see cref="ToolApprovalRequestContent"/>, even for tools that do not require approval.
+    /// </para>
+    /// <para>
+    /// Setting this property to <see langword="true"/> injects an <see cref="NonApprovalRequiredFunctionBypassingChatClient"/>
+    /// decorator above <see cref="FunctionInvokingChatClient"/> in the pipeline. This decorator identifies approval
+    /// requests for non-approval-required tools, removes them from the response, and stores them in the session.
+    /// On the next request, the stored items are automatically re-injected as approved, so the caller only needs
+    /// to handle approval requests for tools that truly require human approval.
+    /// </para>
+    /// <para>
+    /// This option has no effect when <see cref="UseProvidedChatClientAsIs"/> is <see langword="true"/>.
+    /// When using a custom chat client stack, you can add an <see cref="NonApprovalRequiredFunctionBypassingChatClient"/>
+    /// manually via the <see cref="ChatClientBuilderExtensions.UseNonApprovalRequiredFunctionBypassing"/>
+    /// extension method.
+    /// </para>
+    /// </remarks>
+    /// <value>
+    /// Default is <see langword="false"/>.
+    /// </value>
+    [Experimental(DiagnosticIds.Experiments.AgentsAIExperiments)]
+    public bool EnableNonApprovalRequiredFunctionBypassing { get; set; }
+
     /// <summary>
     /// Creates a new instance of <see cref="ChatClientAgentOptions"/> with the same values as this instance.
     /// </summary>
@@ -199,5 +229,6 @@ public sealed class ChatClientAgentOptions
             ThrowOnChatHistoryProviderConflict = this.ThrowOnChatHistoryProviderConflict,
             RequirePerServiceCallChatHistoryPersistence = this.RequirePerServiceCallChatHistoryPersistence,
             EnableMessageInjection = this.EnableMessageInjection,
+            EnableNonApprovalRequiredFunctionBypassing = this.EnableNonApprovalRequiredFunctionBypassing,
         };
 }

dotnet/src/Microsoft.Agents.AI/ChatClient/ChatClientBuilderExtensions.cs

@@ -148,4 +148,35 @@ public static class ChatClientBuilderExtensions
     {
         return builder.Use(innerClient => new MessageInjectingChatClient(innerClient));
     }
+
+    /// <summary>
+    /// Adds an <see cref="NonApprovalRequiredFunctionBypassingChatClient"/> to the chat client pipeline.
+    /// </summary>
+    /// <remarks>
+    /// <para>
+    /// This decorator should be positioned above the <see cref="FunctionInvokingChatClient"/> in the pipeline
+    /// so that it can intercept approval requests for tools that do not require approval. When
+    /// <see cref="FunctionInvokingChatClient"/> converts all function calls to approval requests (because at
+    /// least one tool requires approval), this decorator removes the requests for non-approval-required tools,
+    /// stores them in the session, and automatically re-injects them as approved on the next request.
+    /// </para>
+    /// <para>
+    /// This extension method is intended for use with custom chat client stacks when
+    /// <see cref="ChatClientAgentOptions.UseProvidedChatClientAsIs"/> is <see langword="true"/>.
+    /// When <see cref="ChatClientAgentOptions.UseProvidedChatClientAsIs"/> is <see langword="false"/> (the default),
+    /// the <see cref="ChatClientAgent"/> automatically injects this decorator when
+    /// <see cref="ChatClientAgentOptions.EnableNonApprovalRequiredFunctionBypassing"/> is <see langword="true"/>.
+    /// </para>
+    /// <para>
+    /// This decorator only works within the context of a running <see cref="ChatClientAgent"/> with
+    /// an active session, and will throw an exception if used in any other stack.
+    /// </para>
+    /// </remarks>
+    /// <param name="builder">The <see cref="ChatClientBuilder"/> to add the decorator to.</param>
+    /// <returns>The <paramref name="builder"/> for chaining.</returns>
+    [Experimental(DiagnosticIds.Experiments.AgentsAIExperiments)]
+    public static ChatClientBuilder UseNonApprovalRequiredFunctionBypassing(this ChatClientBuilder builder)
+    {
+        return builder.Use(innerClient => new NonApprovalRequiredFunctionBypassingChatClient(innerClient));
+    }
 }

dotnet/src/Microsoft.Agents.AI/ChatClient/ChatClientExtensions.cs

@@ -53,6 +53,17 @@ public static class ChatClientExtensions
     {
         var chatBuilder = chatClient.AsBuilder();
 
+        // NonApprovalRequiredFunctionBypassingChatClient is registered before FunctionInvokingChatClient so that
+        // it sits above FICC in the pipeline. ChatClientBuilder.Build applies factories in reverse order,
+        // making the first Use() call outermost. By adding this decorator first, the resulting pipeline is:
+        //   NonApprovalRequiredFunctionBypassingChatClient → FunctionInvokingChatClient → ChatHistoryPersistingChatClient → leaf IChatClient
+        // This allows the decorator to intercept FICC's responses and remove approval requests for tools
+        // that don't actually require approval, storing them for automatic re-injection on the next request.
+        if (options?.EnableNonApprovalRequiredFunctionBypassing is true)
+        {
+            chatBuilder.Use(innerClient => new NonApprovalRequiredFunctionBypassingChatClient(innerClient));
+        }
+
         if (chatClient.GetService<FunctionInvokingChatClient>() is null)
         {
             chatBuilder.Use((innerClient, services) =>

dotnet/src/Microsoft.Agents.AI/ChatClient/NonApprovalRequiredFunctionBypassingChatClient.cs

@@ -0,0 +1,285 @@
+// Copyright (c) Microsoft. All rights reserved.
+
+using System;
+using System.Collections.Generic;
+using System.Linq;
+using System.Runtime.CompilerServices;
+using System.Threading;
+using System.Threading.Tasks;
+using Microsoft.Extensions.AI;
+
+namespace Microsoft.Agents.AI;
+
+/// <summary>
+/// A delegating chat client that automatically removes <see cref="ToolApprovalRequestContent"/> for tools
+/// that do not actually require approval, storing auto-approved results in the session for transparent
+/// re-injection on the next request.
+/// </summary>
+/// <remarks>
+/// <para>
+/// <see cref="FunctionInvokingChatClient"/> has an all-or-nothing behavior for approvals: when any tool
+/// in a response is an <see cref="ApprovalRequiredAIFunction"/>, it converts all <see cref="FunctionCallContent"/>
+/// items to <see cref="ToolApprovalRequestContent"/> — even for tools that do not require approval. This
+/// decorator sits above <see cref="FunctionInvokingChatClient"/> in the pipeline and transparently handles
+/// the non-approval-required items so callers only see approval requests for tools that truly need them.
+/// </para>
+/// <para>
+/// On outbound responses, the decorator identifies <see cref="ToolApprovalRequestContent"/> items for tools
+/// that are not wrapped in <see cref="ApprovalRequiredAIFunction"/>, removes them from the response, and
+/// stores them in the session's <see cref="AgentSessionStateBag"/>. On the next inbound request, the stored
+/// items are re-injected as pre-approved <see cref="ToolApprovalResponseContent"/> so that
+/// <see cref="FunctionInvokingChatClient"/> can process them alongside the caller's human-approved responses.
+/// </para>
+/// <para>
+/// This decorator requires an active <see cref="AIAgent.CurrentRunContext"/> with a non-null
+/// <see cref="AgentRunContext.Session"/>. An <see cref="InvalidOperationException"/> is thrown if no
+/// run context or session is available.
+/// </para>
+/// </remarks>
+internal sealed class NonApprovalRequiredFunctionBypassingChatClient : DelegatingChatClient
+{
+    /// <summary>
+    /// The key used in <see cref="AgentSessionStateBag"/> to store pending auto-approved function calls
+    /// between agent runs.
+    /// </summary>
+    internal const string StateBagKey = "_autoApprovedFunctionCalls";
+
+    /// <summary>
+    /// Initializes a new instance of the <see cref="NonApprovalRequiredFunctionBypassingChatClient"/> class.
+    /// </summary>
+    /// <param name="innerClient">The underlying chat client (typically a <see cref="FunctionInvokingChatClient"/>).</param>
+    public NonApprovalRequiredFunctionBypassingChatClient(IChatClient innerClient)
+        : base(innerClient)
+    {
+    }
+
+    /// <inheritdoc/>
+    public override async Task<ChatResponse> GetResponseAsync(
+        IEnumerable<ChatMessage> messages,
+        ChatOptions? options = null,
+        CancellationToken cancellationToken = default)
+    {
+        var session = GetRequiredSession();
+        var autoApprovableNames = this.GetAutoApprovableToolNames(options);
+
+        messages = InjectPendingAutoApprovals(messages, session);
+
+        var response = await base.GetResponseAsync(messages, options, cancellationToken).ConfigureAwait(false);
+
+        RemoveAutoApprovedFromMessages(response.Messages, autoApprovableNames, session);
+
+        return response;
+    }
+
+    /// <inheritdoc/>
+    public override async IAsyncEnumerable<ChatResponseUpdate> GetStreamingResponseAsync(
+        IEnumerable<ChatMessage> messages,
+        ChatOptions? options = null,
+        [EnumeratorCancellation] CancellationToken cancellationToken = default)
+    {
+        var session = GetRequiredSession();
+        var autoApprovableNames = this.GetAutoApprovableToolNames(options);
+
+        messages = InjectPendingAutoApprovals(messages, session);
+        List<ToolApprovalRequestContent>? autoApproved = null;
+
+        try
+        {
+            await foreach (var update in base.GetStreamingResponseAsync(messages, options, cancellationToken).ConfigureAwait(false))
+            {
+                if (FilterUpdateContents(update, autoApprovableNames, ref autoApproved))
+                {
+                    yield return update;
+                }
+            }
+        }
+        finally
+        {
+            if (autoApproved is { Count: > 0 })
+            {
+                session.StateBag.SetValue(StateBagKey, autoApproved, AgentJsonUtilities.DefaultOptions);
+            }
+        }
+    }
+
+    /// <summary>
+    /// Gets the current <see cref="AgentSession"/> from the ambient run context.
+    /// </summary>
+    /// <exception cref="InvalidOperationException">No run context or session is available.</exception>
+    private static AgentSession GetRequiredSession()
+    {
+        var runContext = AIAgent.CurrentRunContext
+            ?? throw new InvalidOperationException(
+                $"{nameof(NonApprovalRequiredFunctionBypassingChatClient)} can only be used within the context of a running AIAgent. " +
+                "Ensure that the chat client is being invoked as part of an AIAgent.RunAsync or AIAgent.RunStreamingAsync call.");
+
+        return runContext.Session
+            ?? throw new InvalidOperationException(
+                $"{nameof(NonApprovalRequiredFunctionBypassingChatClient)} requires a session. " +
+                "Ensure the agent has a resolved session before invoking the chat client.");
+    }
+
+    /// <summary>
+    /// Checks the session for stored auto-approvals from a previous turn and injects them as
+    /// a user message containing <see cref="ToolApprovalResponseContent"/> items appended to the input messages.
+    /// </summary>
+    /// <remarks>
+    /// All stored requests are unconditionally injected as approved responses regardless of whether the
+    /// tool set has changed, because the LLM requires a complete set of tool call responses for a prior turn.
+    /// </remarks>
+    private static IEnumerable<ChatMessage> InjectPendingAutoApprovals(
+        IEnumerable<ChatMessage> messages,
+        AgentSession session)
+    {
+        if (!session.StateBag.TryGetValue<List<ToolApprovalRequestContent>>(
+            StateBagKey,
+            out var pendingRequests,
+            AgentJsonUtilities.DefaultOptions)
+            || pendingRequests is not { Count: > 0 })
+        {
+            return messages;
+        }
+
+        session.StateBag.TryRemoveValue(StateBagKey);
+
+        List<AIContent> approvalResponses = [];
+        foreach (var request in pendingRequests)
+        {
+            approvalResponses.Add(request.CreateResponse(approved: true));
+        }
+
+        var userMessage = new ChatMessage(ChatRole.User, approvalResponses);
+        return messages.Concat([userMessage]);
+    }
+
+    /// <summary>
+    /// Builds a set of tool names that do not require approval and can be auto-approved,
+    /// by checking all available tools from <see cref="ChatOptions.Tools"/> and
+    /// <see cref="FunctionInvokingChatClient.AdditionalTools"/>.
+    /// </summary>
+    private HashSet<string> GetAutoApprovableToolNames(ChatOptions? options)
+    {
+        var ficc = this.GetService<FunctionInvokingChatClient>();
+
+        var allTools = (options?.Tools ?? Enumerable.Empty<AITool>())
+            .Concat(ficc?.AdditionalTools ?? Enumerable.Empty<AITool>());
+
+        return new HashSet<string>(
+            allTools
+                .OfType<AIFunction>()
+                .Where(static f => f.GetService<ApprovalRequiredAIFunction>() is null)
+                .Select(static f => f.Name),
+            StringComparer.Ordinal);
+    }
+
+    /// <summary>
+    /// Determines whether a <see cref="ToolApprovalRequestContent"/> can be auto-approved because
+    /// the underlying tool is not an <see cref="ApprovalRequiredAIFunction"/>.
+    /// </summary>
+    /// <returns>
+    /// <see langword="true"/> if the approval request is for a known tool that does not require approval
+    /// and can be auto-approved; <see langword="false"/> otherwise.
+    /// </returns>
+    private static bool IsAutoApprovable(ToolApprovalRequestContent approval, HashSet<string> autoApprovableNames)
+    {
+        if (approval.ToolCall is not FunctionCallContent fcc)
+        {
+            // Non-function tool calls cannot be auto-approved.
+            return false;
+        }
+
+        // Auto-approve only if the tool is known and explicitly does NOT require approval.
+        // Unknown tools are not in the set and are treated as approval-required (safe default).
+        return autoApprovableNames.Contains(fcc.Name);
+    }
+
+    /// <summary>
+    /// Scans response messages for auto-approvable <see cref="ToolApprovalRequestContent"/> items,
+    /// removes them from the messages, and stores them in the session for the next request.
+    /// </summary>
+    private static void RemoveAutoApprovedFromMessages(
+        IList<ChatMessage> messages,
+        HashSet<string> autoApprovableNames,
+        AgentSession session)
+    {
+        List<ToolApprovalRequestContent>? autoApproved = null;
+
+        foreach (var message in messages)
+        {
+            for (int i = message.Contents.Count - 1; i >= 0; i--)
+            {
+                if (message.Contents[i] is ToolApprovalRequestContent approval
+                    && IsAutoApprovable(approval, autoApprovableNames))
+                {
+                    (autoApproved ??= []).Add(approval);
+                    message.Contents.RemoveAt(i);
+                }
+            }
+        }
+
+        // Remove messages that are now empty after filtering.
+        for (int i = messages.Count - 1; i >= 0; i--)
+        {
+            if (messages[i].Contents.Count == 0)
+            {
+                messages.RemoveAt(i);
+            }
+        }
+
+        if (autoApproved is { Count: > 0 })
+        {
+            session.StateBag.SetValue(StateBagKey, autoApproved, AgentJsonUtilities.DefaultOptions);
+        }
+    }
+
+    /// <summary>
+    /// Filters auto-approvable <see cref="ToolApprovalRequestContent"/> items from a streaming update's
+    /// contents, collecting them for later storage.
+    /// </summary>
+    /// <returns>
+    /// <see langword="true"/> if the update should be yielded (has remaining content or had no
+    /// approval content to begin with); <see langword="false"/> if the update is now empty and
+    /// should be skipped.
+    /// </returns>
+    private static bool FilterUpdateContents(
+        ChatResponseUpdate update,
+        HashSet<string> autoApprovableNames,
+        ref List<ToolApprovalRequestContent>? autoApproved)
+    {
+        bool hasApprovalContent = false;
+        List<AIContent> filteredContents = [];
+        bool removedAny = false;
+
+        for (int i = 0; i < update.Contents.Count; i++)
+        {
+            var content = update.Contents[i];
+
+            if (content is ToolApprovalRequestContent approval)
+            {
+                hasApprovalContent = true;
+
+                if (IsAutoApprovable(approval, autoApprovableNames))
+                {
+                    (autoApproved ??= []).Add(approval);
+                    removedAny = true;
+                }
+                else
+                {
+                    filteredContents.Add(content);
+                }
+            }
+            else
+            {
+                filteredContents.Add(content);
+            }
+        }
+
+        if (removedAny)
+        {
+            update.Contents = filteredContents;
+        }
+
+        // Yield the update unless it was purely auto-approvable approval content (now empty).
+        return update.Contents.Count > 0 || !hasApprovalContent;
+    }
+}

dotnet/src/Microsoft.Agents.AI/Harness/ToolApproval/ToolApprovalAgent.cs

@@ -51,20 +51,22 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
 {
     private readonly ProviderSessionState<ToolApprovalState> _sessionState;
     private readonly JsonSerializerOptions _jsonSerializerOptions;
+    private readonly Func<FunctionCallContent, ValueTask<bool>>[]? _autoApprovalRules;
 
     /// <summary>
     /// Initializes a new instance of the <see cref="ToolApprovalAgent"/> class.
     /// </summary>
     /// <param name="innerAgent">The underlying agent to delegate to.</param>
-    /// <param name="jsonSerializerOptions">
-    /// Optional <see cref="JsonSerializerOptions"/> used for serializing argument values when storing rules
-    /// and for persisting state. When <see langword="null"/>, <see cref="AgentJsonUtilities.DefaultOptions"/> is used.
+    /// <param name="options">
+    /// Optional <see cref="ToolApprovalAgentOptions"/> for configuring serialization and auto-approval rules.
+    /// When <see langword="null"/>, default settings are used.
     /// </param>
     /// <exception cref="ArgumentNullException"><paramref name="innerAgent"/> is <see langword="null"/>.</exception>
-    public ToolApprovalAgent(AIAgent innerAgent, JsonSerializerOptions? jsonSerializerOptions = null)
+    public ToolApprovalAgent(AIAgent innerAgent, ToolApprovalAgentOptions? options = null)
         : base(innerAgent)
     {
-        this._jsonSerializerOptions = jsonSerializerOptions ?? AgentJsonUtilities.DefaultOptions;
+        this._jsonSerializerOptions = options?.JsonSerializerOptions ?? AgentJsonUtilities.DefaultOptions;
+        this._autoApprovalRules = options?.AutoApprovalRules?.ToArray();
         this._sessionState = new ProviderSessionState<ToolApprovalState>(
             _ => new ToolApprovalState(),
             "toolApprovalState",
@@ -79,7 +81,7 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
         CancellationToken cancellationToken = default)
     {
         // Steps 1–2: Unwrap AlwaysApprove wrappers, process any queued approval requests.
-        var (state, callerMessages, nextQueuedItem) = this.PrepareInboundMessages(messages, session);
+        var (state, callerMessages, nextQueuedItem) = await this.PrepareInboundMessagesAsync(messages, session).ConfigureAwait(false);
 
         if (nextQueuedItem is not null)
         {
@@ -98,7 +100,7 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
             var response = await this.InnerAgent.RunAsync(processedMessages, session, options, cancellationToken).ConfigureAwait(false);
 
             // Classify approval requests: auto-approve matching, queue excess, keep first unapproved.
-            bool allAutoApproved = this.ProcessAndQueueOutboundApprovalRequests(response.Messages, state, session);
+            bool allAutoApproved = await this.ProcessAndQueueOutboundApprovalRequestsAsync(response.Messages, state, session).ConfigureAwait(false);
 
             if (!allAutoApproved)
             {
@@ -119,7 +121,7 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
         [EnumeratorCancellation] CancellationToken cancellationToken = default)
     {
         // Steps 1–2: Unwrap AlwaysApprove wrappers, process any queued approval requests.
-        var (state, callerMessages, nextQueuedItem) = this.PrepareInboundMessages(messages, session);
+        var (state, callerMessages, nextQueuedItem) = await this.PrepareInboundMessagesAsync(messages, session).ConfigureAwait(false);
 
         if (nextQueuedItem is not null)
         {
@@ -197,7 +199,7 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
                 yield break;
             }
 
-            // 4. Classify the collected approval requests against standing rules.
+            // 4. Classify the collected approval requests against standing rules and auto-approval rules.
             List<ToolApprovalRequestContent> unapproved = [];
             foreach (var tarc in streamedApprovalRequests)
             {
@@ -206,6 +208,11 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
                     state.CollectedApprovalResponses.Add(
                         tarc.CreateResponse(approved: true, reason: "Auto-approved by standing rule"));
                 }
+                else if (await this.MatchesAutoApprovalRuleAsync(tarc).ConfigureAwait(false))
+                {
+                    state.CollectedApprovalResponses.Add(
+                        tarc.CreateResponse(approved: true, reason: "Auto-approved by auto-approval rule"));
+                }
                 else
                 {
                     unapproved.Add(tarc);
@@ -291,9 +298,9 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
     }
 
     /// <summary>
-    /// Re-evaluates queued approval requests against current rules and auto-approves any that now match.
+    /// Re-evaluates queued approval requests against current rules and auto-approval rules, and auto-approves any that now match.
     /// </summary>
-    private void DrainAutoApprovableFromQueue(ToolApprovalState state)
+    private async ValueTask DrainAutoApprovableFromQueueAsync(ToolApprovalState state)
     {
         for (int i = state.QueuedApprovalRequests.Count - 1; i >= 0; i--)
         {
@@ -303,6 +310,12 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
                     state.QueuedApprovalRequests[i].CreateResponse(approved: true, reason: "Auto-approved by standing rule"));
                 state.QueuedApprovalRequests.RemoveAt(i);
             }
+            else if (await this.MatchesAutoApprovalRuleAsync(state.QueuedApprovalRequests[i]).ConfigureAwait(false))
+            {
+                state.CollectedApprovalResponses.Add(
+                    state.QueuedApprovalRequests[i].CreateResponse(approved: true, reason: "Auto-approved by auto-approval rule"));
+                state.QueuedApprovalRequests.RemoveAt(i);
+            }
         }
     }
 
@@ -318,8 +331,8 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
     /// A tuple of (state, processed caller messages, next queued item or <see langword="null"/> if the queue is resolved).
     /// When the returned item is non-null, the caller should return/yield it without calling the inner agent.
     /// </returns>
-    private (ToolApprovalState State, List<ChatMessage> CallerMessages, ToolApprovalRequestContent? NextQueuedItem)
-        PrepareInboundMessages(IEnumerable<ChatMessage> messages, AgentSession? session)
+    private async ValueTask<(ToolApprovalState State, List<ChatMessage> CallerMessages, ToolApprovalRequestContent? NextQueuedItem)>
+        PrepareInboundMessagesAsync(IEnumerable<ChatMessage> messages, AgentSession? session)
     {
         var state = this._sessionState.GetOrInitializeState(session);
 
@@ -337,7 +350,7 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
 
             // Re-evaluate remaining queued items — the caller may have added new rules
             // (e.g., "always approve this tool") that resolve additional items.
-            this.DrainAutoApprovableFromQueue(state);
+            await this.DrainAutoApprovableFromQueueAsync(state).ConfigureAwait(false);
 
             if (state.QueuedApprovalRequests.Count > 0)
             {
@@ -386,15 +399,18 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
     /// <see langword="true"/> if all TARc items were auto-approved (caller should re-invoke the inner agent);
     /// <see langword="false"/> otherwise.
     /// </returns>
-    private bool ProcessAndQueueOutboundApprovalRequests(
+    private async ValueTask<bool> ProcessAndQueueOutboundApprovalRequestsAsync(
         IList<ChatMessage> responseMessages,
         ToolApprovalState state,
         AgentSession? session)
     {
-        // Pass 1: Scan all response messages and classify each approval request as
-        //         auto-approved (matches a standing rule) or unapproved (needs caller decision).
-        var autoApproved = new List<ToolApprovalRequestContent>();
+        // Pass 1: Scan all response messages and classify each approval request.
+        //         Auto-approved requests (matching a standing rule or auto-approval rule) have their
+        //         responses collected immediately, preserving the original request order, and are
+        //         marked for removal. Unapproved requests are collected for the caller to decide.
+        var toRemove = new HashSet<ToolApprovalRequestContent>();
         var unapproved = new List<ToolApprovalRequestContent>();
+        int autoApprovedCount = 0;
 
         foreach (var message in responseMessages)
         {
@@ -404,7 +420,17 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
                 {
                     if (MatchesRule(tarc, state.Rules, this._jsonSerializerOptions))
                     {
-                        autoApproved.Add(tarc);
+                        state.CollectedApprovalResponses.Add(
+                            tarc.CreateResponse(approved: true, reason: "Auto-approved by standing rule"));
+                        toRemove.Add(tarc);
+                        autoApprovedCount++;
+                    }
+                    else if (await this.MatchesAutoApprovalRuleAsync(tarc).ConfigureAwait(false))
+                    {
+                        state.CollectedApprovalResponses.Add(
+                            tarc.CreateResponse(approved: true, reason: "Auto-approved by auto-approval rule"));
+                        toRemove.Add(tarc);
+                        autoApprovedCount++;
                     }
                     else
                     {
@@ -415,18 +441,12 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
         }
 
         // Nothing to process: no auto-approved items and at most one unapproved (no queueing needed).
-        if (autoApproved.Count == 0 && unapproved.Count <= 1)
+        // No responses were collected above in this case, so state is unmodified and safe to leave.
+        if (autoApprovedCount == 0 && unapproved.Count <= 1)
         {
             return false;
         }
 
-        // Store auto-approved responses for later injection into the inner agent.
-        foreach (var tarc in autoApproved)
-        {
-            state.CollectedApprovalResponses.Add(
-                tarc.CreateResponse(approved: true, reason: "Auto-approved by standing rule"));
-        }
-
         // If every approval request was auto-approved, strip them all and signal the caller
         // to re-invoke the inner agent immediately with the collected responses.
         if (unapproved.Count == 0)
@@ -439,14 +459,10 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
         // Pass 2: Keep only the first unapproved request in the response (for the caller to decide).
         //         Queue the remaining unapproved requests for subsequent one-at-a-time delivery.
         //         Remove all auto-approved and queued items from the response messages.
-        var toRemove = new HashSet<ToolApprovalRequestContent>(autoApproved);
-        if (unapproved.Count > 1)
+        for (int i = 1; i < unapproved.Count; i++)
         {
-            for (int i = 1; i < unapproved.Count; i++)
-            {
-                toRemove.Add(unapproved[i]);
-                state.QueuedApprovalRequests.Add(unapproved[i]);
-            }
+            toRemove.Add(unapproved[i]);
+            state.QueuedApprovalRequests.Add(unapproved[i]);
         }
 
         // Walk messages in reverse and strip marked items.
@@ -663,8 +679,36 @@ public sealed class ToolApprovalAgent : DelegatingAIAgent
     }
 
     /// <summary>
-    /// Compares stored rule arguments against actual function call arguments for an exact match.
+    /// Checks whether a <see cref="ToolApprovalRequestContent"/> is approved by any of the configured
+    /// auto-approval rules (heuristic functions).
     /// </summary>
+    /// <returns>
+    /// <see langword="true"/> if any auto-approval rule returns <see langword="true"/> for the function call;
+    /// <see langword="false"/> if no rules are configured, the request is not a function call, or no rule approves it.
+    /// </returns>
+    private async ValueTask<bool> MatchesAutoApprovalRuleAsync(ToolApprovalRequestContent request)
+    {
+        if (this._autoApprovalRules is not { Length: > 0 })
+        {
+            return false;
+        }
+
+        if (request.ToolCall is not FunctionCallContent functionCall)
+        {
+            return false;
+        }
+
+        foreach (var rule in this._autoApprovalRules)
+        {
+            if (await rule(functionCall).ConfigureAwait(false))
+            {
+                return true;
+            }
+        }
+
+        return false;
+    }
+
     private static bool ArgumentsMatch(IDictionary<string, string> ruleArguments, IDictionary<string, object?>? callArguments, JsonSerializerOptions jsonSerializerOptions)
     {
         if (callArguments is null)

dotnet/src/Microsoft.Agents.AI/Harness/ToolApproval/ToolApprovalAgentBuilderExtensions.cs

@@ -1,7 +1,6 @@
 // Copyright (c) Microsoft. All rights reserved.
 
 using System.Diagnostics.CodeAnalysis;
-using System.Text.Json;
 using Microsoft.Shared.DiagnosticIds;
 using Microsoft.Shared.Diagnostics;
 
@@ -17,9 +16,9 @@ public static class ToolApprovalAgentBuilderExtensions
     /// Adds tool approval middleware to the agent pipeline, enabling "don't ask again" approval behavior.
     /// </summary>
     /// <param name="builder">The <see cref="AIAgentBuilder"/> to which tool approval support will be added.</param>
-    /// <param name="jsonSerializerOptions">
-    /// Optional <see cref="JsonSerializerOptions"/> used for serializing argument values when storing rules
-    /// and for persisting state. When <see langword="null"/>, <see cref="AgentJsonUtilities.DefaultOptions"/> is used.
+    /// <param name="options">
+    /// Optional <see cref="ToolApprovalAgentOptions"/> for configuring serialization and auto-approval rules.
+    /// When <see langword="null"/>, default settings are used.
     /// </param>
     /// <returns>The <see cref="AIAgentBuilder"/> with tool approval middleware added, enabling method chaining.</returns>
     /// <exception cref="System.ArgumentNullException"><paramref name="builder"/> is <see langword="null"/>.</exception>
@@ -32,6 +31,6 @@ public static class ToolApprovalAgentBuilderExtensions
     /// </remarks>
     public static AIAgentBuilder UseToolApproval(
         this AIAgentBuilder builder,
-        JsonSerializerOptions? jsonSerializerOptions = null)
-        => Throw.IfNull(builder).Use(innerAgent => new ToolApprovalAgent(innerAgent, jsonSerializerOptions));
+        ToolApprovalAgentOptions? options = null)
+        => Throw.IfNull(builder).Use(innerAgent => new ToolApprovalAgent(innerAgent, options));
 }

dotnet/src/Microsoft.Agents.AI/Harness/ToolApproval/ToolApprovalAgentOptions.cs

@@ -0,0 +1,45 @@
+// Copyright (c) Microsoft. All rights reserved.
+
+using System;
+using System.Collections.Generic;
+using System.Diagnostics.CodeAnalysis;
+using System.Text.Json;
+using System.Threading.Tasks;
+using Microsoft.Extensions.AI;
+using Microsoft.Shared.DiagnosticIds;
+
+namespace Microsoft.Agents.AI;
+
+/// <summary>
+/// Options for configuring the <see cref="ToolApprovalAgent"/> middleware.
+/// </summary>
+[Experimental(DiagnosticIds.Experiments.AgentsAIExperiments)]
+public class ToolApprovalAgentOptions
+{
+    /// <summary>
+    /// Gets or sets the <see cref="System.Text.Json.JsonSerializerOptions"/> used for serializing argument values
+    /// when storing rules and for persisting state.
+    /// </summary>
+    /// <remarks>
+    /// When <see langword="null"/>, <see cref="AgentJsonUtilities.DefaultOptions"/> is used.
+    /// </remarks>
+    public JsonSerializerOptions? JsonSerializerOptions { get; set; }
+
+    /// <summary>
+    /// Gets or sets a collection of heuristic functions that can automatically approve function calls
+    /// that would otherwise require user approval.
+    /// </summary>
+    /// <remarks>
+    /// <para>
+    /// Each function receives a <see cref="FunctionCallContent"/> representing the tool call that requires approval
+    /// and returns a <see cref="ValueTask{Boolean}"/> that resolves to <see langword="true"/> to auto-approve
+    /// the call, or <see langword="false"/> to continue evaluating the next rule.
+    /// </para>
+    /// <para>
+    /// Auto-approval rules are evaluated after standing rules (derived from prior user approvals) but before
+    /// prompting the user. Rules are evaluated in order; the first rule returning <see langword="true"/>
+    /// causes the function call to be auto-approved.
+    /// </para>
+    /// </remarks>
+    public IEnumerable<Func<FunctionCallContent, ValueTask<bool>>>? AutoApprovalRules { get; set; }
+}

dotnet/tests/Microsoft.Agents.AI.Harness.UnitTests/HarnessAgentTests.cs

@@ -644,6 +644,51 @@ public class HarnessAgentTests
         Assert.Null(agent.GetService<ToolApprovalAgent>());
     }
 
+    /// <summary>
+    /// Verify that ToolApprovalAgentOptions auto-approval rules are passed through and actually used.
+    /// </summary>
+    [Fact]
+    public async Task ToolApproval_AutoApprovalRulesAreAppliedAsync()
+    {
+        // Arrange — inner client returns an approval request on first call, then final response on second.
+        var callCount = 0;
+        var approvalRequest = new ToolApprovalRequestContent("req1", new FunctionCallContent("call1", "ReadTool"));
+
+        var mockClient = new Mock<IChatClient>();
+        mockClient
+            .Setup(c => c.GetResponseAsync(
+                It.IsAny<IEnumerable<ChatMessage>>(),
+                It.IsAny<ChatOptions>(),
+                It.IsAny<CancellationToken>()))
+            .ReturnsAsync(() =>
+            {
+                callCount++;
+                if (callCount == 1)
+                {
+                    return new ChatResponse(new ChatMessage(ChatRole.Assistant, [approvalRequest]));
+                }
+
+                return new ChatResponse(new ChatMessage(ChatRole.Assistant, "Done"));
+            });
+
+        var options = CreateAllDisabledOptions();
+        options.DisableToolApproval = false;
+        options.ToolApprovalAgentOptions = new ToolApprovalAgentOptions
+        {
+            AutoApprovalRules = [fcc => new ValueTask<bool>(fcc.Name == "ReadTool")]
+        };
+
+        var agent = new HarnessAgent(mockClient.Object, TestMaxContextWindowTokens, TestMaxOutputTokens, options);
+        var session = await agent.CreateSessionAsync();
+
+        // Act
+        var response = await agent.RunAsync([new ChatMessage(ChatRole.User, "Hi")], session);
+
+        // Assert — the auto-approval rule approved the request, so we get "Done" (not an approval request)
+        Assert.Equal(2, callCount);
+        Assert.Equal("Done", response.Text);
+    }
+
     #endregion
 
     #region Feature: OpenTelemetry

dotnet/tests/Microsoft.Agents.AI.UnitTests/ChatClient/ChatClientAgentOptionsTests.cs

@@ -134,6 +134,7 @@ public class ChatClientAgentOptionsTests
             ClearOnChatHistoryProviderConflict = false,
             WarnOnChatHistoryProviderConflict = false,
             ThrowOnChatHistoryProviderConflict = false,
+            EnableNonApprovalRequiredFunctionBypassing = true,
         };
 
         // Act
@@ -150,6 +151,7 @@ public class ChatClientAgentOptionsTests
         Assert.Equal(original.ClearOnChatHistoryProviderConflict, clone.ClearOnChatHistoryProviderConflict);
         Assert.Equal(original.WarnOnChatHistoryProviderConflict, clone.WarnOnChatHistoryProviderConflict);
         Assert.Equal(original.ThrowOnChatHistoryProviderConflict, clone.ThrowOnChatHistoryProviderConflict);
+        Assert.Equal(original.EnableNonApprovalRequiredFunctionBypassing, clone.EnableNonApprovalRequiredFunctionBypassing);
 
         // ChatOptions should be cloned, not the same reference
         Assert.NotSame(original.ChatOptions, clone.ChatOptions);

dotnet/tests/Microsoft.Agents.AI.UnitTests/ChatClient/NonApprovalRequiredFunctionBypassingChatClientTests.cs

@@ -0,0 +1,574 @@
+// Copyright (c) Microsoft. All rights reserved.
+
+using System;
+using System.Collections.Generic;
+using System.Linq;
+using System.Threading;
+using System.Threading.Tasks;
+using Microsoft.Extensions.AI;
+using Moq;
+
+namespace Microsoft.Agents.AI.UnitTests;
+
+public class NonApprovalRequiredFunctionBypassingChatClientTests
+{
+    #region GetResponseAsync Tests
+
+    [Fact]
+    public async Task GetResponseAsync_NoApprovalContent_PassesThroughUnchangedAsync()
+    {
+        // Arrange
+        var innerClient = CreateMockChatClient((_, _, _) =>
+            Task.FromResult(new ChatResponse([new ChatMessage(ChatRole.Assistant, "Hello")])));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var session = new ChatClientAgentSession();
+
+        // Act
+        var response = await RunWithAgentContextAsync(decorator, session);
+
+        // Assert
+        Assert.Single(response.Messages);
+        Assert.Equal("Hello", response.Messages[0].Text);
+        Assert.Equal(0, session.StateBag.Count);
+    }
+
+    [Fact]
+    public async Task GetResponseAsync_AllToolsRequireApproval_PassesThroughUnchangedAsync()
+    {
+        // Arrange
+        var approvalTool = new ApprovalRequiredAIFunction(AIFunctionFactory.Create(() => "result", "approvalTool"));
+        var fcc = new FunctionCallContent("call1", "approvalTool");
+        var approval = new ToolApprovalRequestContent("req1", fcc);
+
+        var innerClient = CreateMockChatClient((_, _, _) =>
+            Task.FromResult(new ChatResponse([new ChatMessage(ChatRole.Assistant, [approval])])));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var session = new ChatClientAgentSession();
+        var options = new ChatOptions { Tools = [approvalTool] };
+
+        // Act
+        var response = await RunWithAgentContextAsync(decorator, session, options);
+
+        // Assert — approval request should remain
+        Assert.Single(response.Messages);
+        var contents = response.Messages[0].Contents;
+        Assert.Single(contents);
+        Assert.IsType<ToolApprovalRequestContent>(contents[0]);
+        Assert.Equal(0, session.StateBag.Count);
+    }
+
+    [Fact]
+    public async Task GetResponseAsync_MixedApproval_RemovesNonApprovalItemsAsync()
+    {
+        // Arrange
+        var normalTool = AIFunctionFactory.Create(() => "result", "normalTool");
+        var approvalTool = new ApprovalRequiredAIFunction(AIFunctionFactory.Create(() => "result", "approvalTool"));
+
+        var fccNormal = new FunctionCallContent("call1", "normalTool");
+        var fccApproval = new FunctionCallContent("call2", "approvalTool");
+        var approvalNormal = new ToolApprovalRequestContent("req1", fccNormal);
+        var approvalRequired = new ToolApprovalRequestContent("req2", fccApproval);
+
+        var innerClient = CreateMockChatClient((_, _, _) =>
+            Task.FromResult(new ChatResponse([
+                new ChatMessage(ChatRole.Assistant, [approvalNormal, approvalRequired])
+            ])));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var session = new ChatClientAgentSession();
+        var options = new ChatOptions { Tools = [normalTool, approvalTool] };
+
+        // Act
+        var response = await RunWithAgentContextAsync(decorator, session, options);
+
+        // Assert — only the approval-required item remains in the response
+        Assert.Single(response.Messages);
+        var contents = response.Messages[0].Contents;
+        Assert.Single(contents);
+        var remainingApproval = Assert.IsType<ToolApprovalRequestContent>(contents[0]);
+        Assert.Equal("req2", remainingApproval.RequestId);
+    }
+
+    [Fact]
+    public async Task GetResponseAsync_MixedApproval_StoresAutoApprovedInSessionAsync()
+    {
+        // Arrange
+        var normalTool = AIFunctionFactory.Create(() => "result", "normalTool");
+        var approvalTool = new ApprovalRequiredAIFunction(AIFunctionFactory.Create(() => "result", "approvalTool"));
+
+        var fccNormal = new FunctionCallContent("call1", "normalTool");
+        var fccApproval = new FunctionCallContent("call2", "approvalTool");
+        var approvalNormal = new ToolApprovalRequestContent("req1", fccNormal);
+        var approvalRequired = new ToolApprovalRequestContent("req2", fccApproval);
+
+        var innerClient = CreateMockChatClient((_, _, _) =>
+            Task.FromResult(new ChatResponse([
+                new ChatMessage(ChatRole.Assistant, [approvalNormal, approvalRequired])
+            ])));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var session = new ChatClientAgentSession();
+        var options = new ChatOptions { Tools = [normalTool, approvalTool] };
+
+        // Act
+        await RunWithAgentContextAsync(decorator, session, options);
+
+        // Assert — the auto-approved item should be stored in the session
+        Assert.True(session.StateBag.TryGetValue<List<ToolApprovalRequestContent>>(
+            NonApprovalRequiredFunctionBypassingChatClient.StateBagKey, out var stored, AgentJsonUtilities.DefaultOptions));
+        Assert.NotNull(stored);
+        Assert.Single(stored!);
+        Assert.Equal("req1", stored![0].RequestId);
+    }
+
+    [Fact]
+    public async Task GetResponseAsync_AllNonApproval_RemovesAllApprovalsAndRemovesEmptyMessageAsync()
+    {
+        // Arrange
+        var normalTool = AIFunctionFactory.Create(() => "result", "normalTool");
+
+        var fccNormal = new FunctionCallContent("call1", "normalTool");
+        var approvalNormal = new ToolApprovalRequestContent("req1", fccNormal);
+
+        var innerClient = CreateMockChatClient((_, _, _) =>
+            Task.FromResult(new ChatResponse([
+                new ChatMessage(ChatRole.Assistant, [approvalNormal])
+            ])));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var session = new ChatClientAgentSession();
+        var options = new ChatOptions { Tools = [normalTool] };
+
+        // Act
+        var response = await RunWithAgentContextAsync(decorator, session, options);
+
+        // Assert — the message should be removed since it's now empty
+        Assert.Empty(response.Messages);
+    }
+
+    [Fact]
+    public async Task GetResponseAsync_NextRequest_InjectsStoredAutoApprovalsAsync()
+    {
+        // Arrange
+        var fccNormal = new FunctionCallContent("call1", "normalTool");
+        var storedApproval = new ToolApprovalRequestContent("req1", fccNormal);
+
+        var session = new ChatClientAgentSession();
+        session.StateBag.SetValue(
+            NonApprovalRequiredFunctionBypassingChatClient.StateBagKey,
+            new List<ToolApprovalRequestContent> { storedApproval },
+            AgentJsonUtilities.DefaultOptions);
+
+        IEnumerable<ChatMessage>? capturedMessages = null;
+        var innerClient = CreateMockChatClient((messages, _, _) =>
+        {
+            capturedMessages = messages.ToList();
+            return Task.FromResult(new ChatResponse([new ChatMessage(ChatRole.Assistant, "Done")]));
+        });
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var options = new ChatOptions { Tools = [AIFunctionFactory.Create(() => "result", "normalTool")] };
+
+        // Act
+        await RunWithAgentContextAsync(decorator, session, options);
+
+        // Assert — the inner client should receive injected messages
+        Assert.NotNull(capturedMessages);
+        var messagesList = capturedMessages!.ToList();
+
+        // Original user message + user message with approved responses.
+        Assert.Equal(2, messagesList.Count);
+        Assert.Equal(ChatRole.User, messagesList[0].Role);
+
+        // User message with the auto-approved ToolApprovalResponseContent
+        Assert.Equal(ChatRole.User, messagesList[1].Role);
+        var userContent = messagesList[1].Contents.OfType<ToolApprovalResponseContent>().ToList();
+        Assert.Single(userContent);
+        Assert.Equal("req1", userContent[0].RequestId);
+        Assert.True(userContent[0].Approved);
+    }
+
+    [Fact]
+    public async Task GetResponseAsync_NextRequest_ClearsStoredAfterInjectionAsync()
+    {
+        // Arrange
+        var fccNormal = new FunctionCallContent("call1", "normalTool");
+        var storedApproval = new ToolApprovalRequestContent("req1", fccNormal);
+
+        var session = new ChatClientAgentSession();
+        session.StateBag.SetValue(
+            NonApprovalRequiredFunctionBypassingChatClient.StateBagKey,
+            new List<ToolApprovalRequestContent> { storedApproval },
+            AgentJsonUtilities.DefaultOptions);
+
+        var innerClient = CreateMockChatClient((_, _, _) =>
+            Task.FromResult(new ChatResponse([new ChatMessage(ChatRole.Assistant, "Done")])));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var options = new ChatOptions { Tools = [AIFunctionFactory.Create(() => "result", "normalTool")] };
+
+        // Act
+        await RunWithAgentContextAsync(decorator, session, options);
+
+        // Assert — the stored data should be cleared after successful injection
+        Assert.False(session.StateBag.TryGetValue<List<ToolApprovalRequestContent>>(
+            NonApprovalRequiredFunctionBypassingChatClient.StateBagKey, out _, AgentJsonUtilities.DefaultOptions));
+    }
+
+    [Fact]
+    public async Task GetResponseAsync_UnknownTool_TreatedAsApprovalRequiredAsync()
+    {
+        // Arrange — tool is not in ChatOptions.Tools
+        var fccUnknown = new FunctionCallContent("call1", "unknownTool");
+        var approvalUnknown = new ToolApprovalRequestContent("req1", fccUnknown);
+
+        var innerClient = CreateMockChatClient((_, _, _) =>
+            Task.FromResult(new ChatResponse([
+                new ChatMessage(ChatRole.Assistant, [approvalUnknown])
+            ])));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var session = new ChatClientAgentSession();
+        var options = new ChatOptions { Tools = [] };
+
+        // Act
+        var response = await RunWithAgentContextAsync(decorator, session, options);
+
+        // Assert — unknown tool should NOT be auto-approved
+        Assert.Single(response.Messages);
+        Assert.Single(response.Messages[0].Contents);
+        Assert.IsType<ToolApprovalRequestContent>(response.Messages[0].Contents[0]);
+        Assert.Equal(0, session.StateBag.Count);
+    }
+
+    [Fact]
+    public async Task GetResponseAsync_StoredRequestToolSetChanged_StillInjectsAsApprovedAsync()
+    {
+        // Arrange — tool was previously non-approval-required but is now wrapped in ApprovalRequiredAIFunction.
+        // The LLM still requires a complete set of responses, so we inject unconditionally.
+        var fccTool = new FunctionCallContent("call1", "changingTool");
+        var storedApproval = new ToolApprovalRequestContent("req1", fccTool);
+
+        var session = new ChatClientAgentSession();
+        session.StateBag.SetValue(
+            NonApprovalRequiredFunctionBypassingChatClient.StateBagKey,
+            new List<ToolApprovalRequestContent> { storedApproval },
+            AgentJsonUtilities.DefaultOptions);
+
+        IEnumerable<ChatMessage>? capturedMessages = null;
+        var innerClient = CreateMockChatClient((messages, _, _) =>
+        {
+            capturedMessages = messages.ToList();
+            return Task.FromResult(new ChatResponse([new ChatMessage(ChatRole.Assistant, "Done")]));
+        });
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+
+        // The tool is now wrapped in ApprovalRequiredAIFunction — but we still inject unconditionally
+        var approvalTool = new ApprovalRequiredAIFunction(AIFunctionFactory.Create(() => "result", "changingTool"));
+        var options = new ChatOptions { Tools = [approvalTool] };
+
+        // Act
+        await RunWithAgentContextAsync(decorator, session, options);
+
+        // Assert — the stored request should still be injected as approved
+        Assert.NotNull(capturedMessages);
+        var messagesList = capturedMessages!.ToList();
+        Assert.Equal(2, messagesList.Count);
+        var userContent = messagesList[1].Contents.OfType<ToolApprovalResponseContent>().ToList();
+        Assert.Single(userContent);
+        Assert.Equal("req1", userContent[0].RequestId);
+        Assert.True(userContent[0].Approved);
+
+        // Session should be cleared
+        Assert.False(session.StateBag.TryGetValue<List<ToolApprovalRequestContent>>(
+            NonApprovalRequiredFunctionBypassingChatClient.StateBagKey, out _, AgentJsonUtilities.DefaultOptions));
+    }
+
+    #endregion
+
+    #region GetStreamingResponseAsync Tests
+
+    [Fact]
+    public async Task GetStreamingResponseAsync_NoApprovalContent_PassesThroughUnchangedAsync()
+    {
+        // Arrange
+        var innerClient = CreateMockStreamingChatClient((_, _, _) =>
+            ToAsyncEnumerableAsync(
+                new ChatResponseUpdate(ChatRole.Assistant, "Hello")));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var session = new ChatClientAgentSession();
+
+        // Act
+        var updates = new List<ChatResponseUpdate>();
+        await RunStreamingWithAgentContextAsync(decorator, session, updates);
+
+        // Assert
+        Assert.Single(updates);
+        Assert.Equal("Hello", updates[0].Text);
+        Assert.Equal(0, session.StateBag.Count);
+    }
+
+    [Fact]
+    public async Task GetStreamingResponseAsync_MixedApproval_FiltersNonApprovalItemsAsync()
+    {
+        // Arrange
+        var normalTool = AIFunctionFactory.Create(() => "result", "normalTool");
+        var approvalTool = new ApprovalRequiredAIFunction(AIFunctionFactory.Create(() => "result", "approvalTool"));
+
+        var fccNormal = new FunctionCallContent("call1", "normalTool");
+        var fccApproval = new FunctionCallContent("call2", "approvalTool");
+        var approvalNormal = new ToolApprovalRequestContent("req1", fccNormal);
+        var approvalRequired = new ToolApprovalRequestContent("req2", fccApproval);
+
+        var innerClient = CreateMockStreamingChatClient((_, _, _) =>
+            ToAsyncEnumerableAsync(
+                new ChatResponseUpdate(ChatRole.Assistant, "text"),
+                new ChatResponseUpdate { Contents = [approvalNormal, approvalRequired] }));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var session = new ChatClientAgentSession();
+        var options = new ChatOptions { Tools = [normalTool, approvalTool] };
+
+        // Act
+        var updates = new List<ChatResponseUpdate>();
+        await RunStreamingWithAgentContextAsync(decorator, session, updates, options);
+
+        // Assert — text update + filtered approval update
+        Assert.Equal(2, updates.Count);
+        Assert.Equal("text", updates[0].Text);
+
+        // Second update should only have the approval-required item
+        var approvalContents = updates[1].Contents.OfType<ToolApprovalRequestContent>().ToList();
+        Assert.Single(approvalContents);
+        Assert.Equal("req2", approvalContents[0].RequestId);
+    }
+
+    [Fact]
+    public async Task GetStreamingResponseAsync_MixedApproval_StoresAutoApprovedInSessionAsync()
+    {
+        // Arrange
+        var normalTool = AIFunctionFactory.Create(() => "result", "normalTool");
+        var approvalTool = new ApprovalRequiredAIFunction(AIFunctionFactory.Create(() => "result", "approvalTool"));
+
+        var fccNormal = new FunctionCallContent("call1", "normalTool");
+        var fccApproval = new FunctionCallContent("call2", "approvalTool");
+        var approvalNormal = new ToolApprovalRequestContent("req1", fccNormal);
+        var approvalRequired = new ToolApprovalRequestContent("req2", fccApproval);
+
+        var innerClient = CreateMockStreamingChatClient((_, _, _) =>
+            ToAsyncEnumerableAsync(
+                new ChatResponseUpdate { Contents = [approvalNormal, approvalRequired] }));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var session = new ChatClientAgentSession();
+        var options = new ChatOptions { Tools = [normalTool, approvalTool] };
+
+        // Act
+        var updates = new List<ChatResponseUpdate>();
+        await RunStreamingWithAgentContextAsync(decorator, session, updates, options);
+
+        // Assert — the auto-approved item should be stored in the session
+        Assert.True(session.StateBag.TryGetValue<List<ToolApprovalRequestContent>>(
+            NonApprovalRequiredFunctionBypassingChatClient.StateBagKey, out var stored, AgentJsonUtilities.DefaultOptions));
+        Assert.NotNull(stored);
+        Assert.Single(stored!);
+        Assert.Equal("req1", stored![0].RequestId);
+    }
+
+    [Fact]
+    public async Task GetStreamingResponseAsync_AllNonApproval_SkipsEmptyUpdateAsync()
+    {
+        // Arrange
+        var normalTool = AIFunctionFactory.Create(() => "result", "normalTool");
+
+        var fccNormal = new FunctionCallContent("call1", "normalTool");
+        var approvalNormal = new ToolApprovalRequestContent("req1", fccNormal);
+
+        var innerClient = CreateMockStreamingChatClient((_, _, _) =>
+            ToAsyncEnumerableAsync(
+                new ChatResponseUpdate(ChatRole.Assistant, "text"),
+                new ChatResponseUpdate { Contents = [approvalNormal] }));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+        var session = new ChatClientAgentSession();
+        var options = new ChatOptions { Tools = [normalTool] };
+
+        // Act
+        var updates = new List<ChatResponseUpdate>();
+        await RunStreamingWithAgentContextAsync(decorator, session, updates, options);
+
+        // Assert — the approval update should be skipped entirely
+        Assert.Single(updates);
+        Assert.Equal("text", updates[0].Text);
+    }
+
+    #endregion
+
+    #region Error Handling Tests
+
+    [Fact]
+    public async Task GetResponseAsync_NoRunContext_ThrowsInvalidOperationExceptionAsync()
+    {
+        // Arrange
+        var innerClient = CreateMockChatClient((_, _, _) =>
+            Task.FromResult(new ChatResponse([new ChatMessage(ChatRole.Assistant, "response")])));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+
+        // Act & Assert — calling directly without agent context
+        await Assert.ThrowsAsync<InvalidOperationException>(
+            () => decorator.GetResponseAsync([new ChatMessage(ChatRole.User, "test")]));
+    }
+
+    [Fact]
+    public async Task GetResponseAsync_NoSession_ThrowsInvalidOperationExceptionAsync()
+    {
+        // Arrange
+        var innerClient = CreateMockChatClient((_, _, _) =>
+            Task.FromResult(new ChatResponse([new ChatMessage(ChatRole.Assistant, "response")])));
+
+        var decorator = new NonApprovalRequiredFunctionBypassingChatClient(innerClient);
+
+        // Act & Assert — run with null session
+        await Assert.ThrowsAsync<InvalidOperationException>(
+            () => RunWithAgentContextAsync(decorator, session: null!));
+    }
+
+    #endregion
+
+    #region Builder Extension Tests
+
+    [Fact]
+    public void UseNonApprovalRequiredFunctionBypassing_AddsDecoratorToPipeline()
+    {
+        // Arrange
+        var innerClient = new Mock<IChatClient>().Object;
+
+        // Act
+        var pipeline = innerClient.AsBuilder()
+            .UseNonApprovalRequiredFunctionBypassing()
+            .Build();
+
+        // Assert
+        Assert.NotNull(pipeline.GetService<NonApprovalRequiredFunctionBypassingChatClient>());
+    }
+
+    [Fact]
+    public void WithDefaultAgentMiddleware_EnableNonApprovalRequiredFunctionBypassing_InjectsDecorator()
+    {
+        // Arrange
+        var innerClient = new Mock<IChatClient>().Object;
+        var options = new ChatClientAgentOptions { EnableNonApprovalRequiredFunctionBypassing = true };
+
+        // Act
+        var pipeline = innerClient.WithDefaultAgentMiddleware(options);
+
+        // Assert
+        Assert.NotNull(pipeline.GetService<NonApprovalRequiredFunctionBypassingChatClient>());
+    }
+
+    [Fact]
+    public void WithDefaultAgentMiddleware_EnableNonApprovalRequiredFunctionBypassingFalse_DoesNotInjectDecorator()
+    {
+        // Arrange
+        var innerClient = new Mock<IChatClient>().Object;
+        var options = new ChatClientAgentOptions { EnableNonApprovalRequiredFunctionBypassing = false };
+
+        // Act
+        var pipeline = innerClient.WithDefaultAgentMiddleware(options);
+
+        // Assert
+        Assert.Null(pipeline.GetService<NonApprovalRequiredFunctionBypassingChatClient>());
+    }
+
+    #endregion
+
+    #region Helpers
+
+    private static async Task<ChatResponse> RunWithAgentContextAsync(
+        NonApprovalRequiredFunctionBypassingChatClient decorator,
+        AgentSession? session,
+        ChatOptions? options = null)
+    {
+        ChatResponse? capturedResponse = null;
+
+        var agent = new TestAIAgent
+        {
+            RunAsyncFunc = async (messages, agentSession, agentOptions, ct) =>
+            {
+                capturedResponse = await decorator.GetResponseAsync(messages, options, ct);
+                return new AgentResponse(capturedResponse);
+            }
+        };
+
+        await agent.RunAsync([new ChatMessage(ChatRole.User, "Hello")], session);
+        return capturedResponse!;
+    }
+
+    private static Task<ChatResponse> RunWithAgentContextAsync(
+        NonApprovalRequiredFunctionBypassingChatClient decorator,
+        AgentSession session)
+        => RunWithAgentContextAsync(decorator, session, options: null);
+
+    private static async Task RunStreamingWithAgentContextAsync(
+        NonApprovalRequiredFunctionBypassingChatClient decorator,
+        AgentSession session,
+        List<ChatResponseUpdate> updates,
+        ChatOptions? options = null)
+    {
+        var agent = new TestAIAgent
+        {
+            RunAsyncFunc = async (messages, agentSession, agentOptions, ct) =>
+            {
+                await foreach (var update in decorator.GetStreamingResponseAsync(messages, options, ct))
+                {
+                    updates.Add(update);
+                }
+
+                return new AgentResponse([new ChatMessage(ChatRole.Assistant, "done")]);
+            }
+        };
+
+        await agent.RunAsync([new ChatMessage(ChatRole.User, "Hello")], session);
+    }
+
+    private static IChatClient CreateMockChatClient(
+        Func<IEnumerable<ChatMessage>, ChatOptions?, CancellationToken, Task<ChatResponse>> onGetResponse)
+    {
+        var mock = new Mock<IChatClient>();
+        mock.Setup(c => c.GetResponseAsync(
+                It.IsAny<IEnumerable<ChatMessage>>(),
+                It.IsAny<ChatOptions?>(),
+                It.IsAny<CancellationToken>()))
+            .Returns((IEnumerable<ChatMessage> m, ChatOptions? o, CancellationToken ct) => onGetResponse(m, o, ct));
+        return mock.Object;
+    }
+
+    private static IChatClient CreateMockStreamingChatClient(
+        Func<IEnumerable<ChatMessage>, ChatOptions?, CancellationToken, IAsyncEnumerable<ChatResponseUpdate>> onGetStreamingResponse)
+    {
+        var mock = new Mock<IChatClient>();
+        mock.Setup(c => c.GetStreamingResponseAsync(
+                It.IsAny<IEnumerable<ChatMessage>>(),
+                It.IsAny<ChatOptions?>(),
+                It.IsAny<CancellationToken>()))
+            .Returns((IEnumerable<ChatMessage> m, ChatOptions? o, CancellationToken ct) => onGetStreamingResponse(m, o, ct));
+        return mock.Object;
+    }
+
+    private static async IAsyncEnumerable<ChatResponseUpdate> ToAsyncEnumerableAsync(params ChatResponseUpdate[] updates)
+    {
+        foreach (var update in updates)
+        {
+            yield return update;
+        }
+
+        await Task.CompletedTask;
+    }
+
+    #endregion
+}

dotnet/tests/Microsoft.Agents.AI.UnitTests/Harness/ToolApproval/ToolApprovalAgentBuilderExtensionsTests.cs

@@ -59,15 +59,15 @@ public class ToolApprovalAgentBuilderExtensionsTests
     /// Verify that UseToolApproval with custom JsonSerializerOptions works correctly.
     /// </summary>
     [Fact]
-    public void UseToolApproval_WithCustomJsonSerializerOptions_ReturnsToolApprovalAgent()
+    public void UseToolApproval_WithCustomOptions_ReturnsToolApprovalAgent()
     {
         // Arrange
         var mockAgent = new Mock<AIAgent>();
         var builder = new AIAgentBuilder(mockAgent.Object);
-        var options = new JsonSerializerOptions();
+        var options = new ToolApprovalAgentOptions { JsonSerializerOptions = new JsonSerializerOptions() };
 
         // Act
-        var result = builder.UseToolApproval(jsonSerializerOptions: options).Build();
+        var result = builder.UseToolApproval(options: options).Build();
 
         // Assert
         Assert.IsType<ToolApprovalAgent>(result);

dotnet/tests/Microsoft.Agents.AI.UnitTests/Harness/ToolApproval/ToolApprovalAgentTests.cs

@@ -47,14 +47,14 @@ public class ToolApprovalAgentTests
     }
 
     /// <summary>
-    /// Verify that constructor accepts custom JsonSerializerOptions.
+    /// Verify that constructor accepts custom options.
     /// </summary>
     [Fact]
-    public void Constructor_CustomJsonSerializerOptions_CreatesInstanceAsync()
+    public void Constructor_CustomOptions_CreatesInstance()
     {
         // Arrange
         var innerAgent = new Mock<AIAgent>().Object;
-        var options = new JsonSerializerOptions();
+        var options = new ToolApprovalAgentOptions { JsonSerializerOptions = new JsonSerializerOptions() };
 
         // Act
         var agent = new ToolApprovalAgent(innerAgent, options);
@@ -1535,4 +1535,311 @@ public class ToolApprovalAgentTests
     }
 
     #endregion
+
+    #region Auto-Approval Rules (Heuristics)
+
+    /// <summary>
+    /// Verify that an auto-approval rule can approve a function call that would otherwise need user approval.
+    /// </summary>
+    [Fact]
+    public async Task RunAsync_AutoApprovalRule_ApprovesMatchingToolAsync()
+    {
+        // Arrange
+        var session = new ChatClientAgentSession();
+        var approvalRequest = new ToolApprovalRequestContent("req1", new FunctionCallContent("call1", "ReadTool"));
+
+        // Inner agent: first call returns approval request, second returns final response.
+        var callCount = 0;
+        var innerAgent = new Mock<AIAgent>();
+        innerAgent
+            .Protected()
+            .Setup<Task<AgentResponse>>("RunCoreAsync",
+                ItExpr.IsAny<IEnumerable<ChatMessage>>(),
+                ItExpr.IsAny<AgentSession?>(),
+                ItExpr.IsAny<AgentRunOptions?>(),
+                ItExpr.IsAny<CancellationToken>())
+            .ReturnsAsync(() =>
+            {
+                callCount++;
+                if (callCount == 1)
+                {
+                    return new AgentResponse([new ChatMessage(ChatRole.Assistant, [approvalRequest])]);
+                }
+
+                return new AgentResponse([new ChatMessage(ChatRole.Assistant, "Done")]);
+            });
+
+        var options = new ToolApprovalAgentOptions
+        {
+            AutoApprovalRules = [fcc => new ValueTask<bool>(fcc.Name == "ReadTool")]
+        };
+        var agent = new ToolApprovalAgent(innerAgent.Object, options);
+
+        // Act
+        var response = await agent.RunAsync(
+            [new ChatMessage(ChatRole.User, "Hi")],
+            session);
+
+        // Assert — the approval request was auto-approved, inner agent called twice
+        Assert.Equal(2, callCount);
+        Assert.Equal("Done", response.Text);
+    }
+
+    /// <summary>
+    /// Verify that when auto-approval rule does not match, request is surfaced to the caller.
+    /// </summary>
+    [Fact]
+    public async Task RunAsync_AutoApprovalRule_DoesNotMatchSurfacesToCallerAsync()
+    {
+        // Arrange
+        var session = new ChatClientAgentSession();
+        var approvalRequest = new ToolApprovalRequestContent("req1", new FunctionCallContent("call1", "DangerousTool"));
+
+        var innerAgent = CreateMockAgent(new AgentResponse([new ChatMessage(ChatRole.Assistant, [approvalRequest])]));
+
+        var options = new ToolApprovalAgentOptions
+        {
+            AutoApprovalRules = [fcc => new ValueTask<bool>(fcc.Name == "ReadTool")]  // Only approves ReadTool
+        };
+        var agent = new ToolApprovalAgent(innerAgent.Object, options);
+
+        // Act
+        var response = await agent.RunAsync(
+            [new ChatMessage(ChatRole.User, "Hi")],
+            session);
+
+        // Assert — request surfaced to caller since heuristic doesn't match
+        var requests = response.Messages.SelectMany(m => m.Contents).OfType<ToolApprovalRequestContent>().ToList();
+        Assert.Single(requests);
+        Assert.Equal("DangerousTool", ((FunctionCallContent)requests[0].ToolCall).Name);
+    }
+
+    /// <summary>
+    /// Verify that multiple auto-approval rules are evaluated in order; first match wins.
+    /// </summary>
+    [Fact]
+    public async Task RunAsync_MultipleAutoApprovalRules_FirstMatchWinsAsync()
+    {
+        // Arrange
+        var session = new ChatClientAgentSession();
+        var approvalRequest = new ToolApprovalRequestContent("req1", new FunctionCallContent("call1", "SpecialTool"));
+
+        var callCount = 0;
+        var innerAgent = new Mock<AIAgent>();
+        innerAgent
+            .Protected()
+            .Setup<Task<AgentResponse>>("RunCoreAsync",
+                ItExpr.IsAny<IEnumerable<ChatMessage>>(),
+                ItExpr.IsAny<AgentSession?>(),
+                ItExpr.IsAny<AgentRunOptions?>(),
+                ItExpr.IsAny<CancellationToken>())
+            .ReturnsAsync(() =>
+            {
+                callCount++;
+                if (callCount == 1)
+                {
+                    return new AgentResponse([new ChatMessage(ChatRole.Assistant, [approvalRequest])]);
+                }
+
+                return new AgentResponse([new ChatMessage(ChatRole.Assistant, "Done")]);
+            });
+
+        var rule1Called = false;
+        var rule2Called = false;
+        var options = new ToolApprovalAgentOptions
+        {
+            AutoApprovalRules =
+            [
+                fcc => { rule1Called = true; return new ValueTask<bool>(fcc.Name == "SpecialTool"); },
+                fcc => { rule2Called = true; return new ValueTask<bool>(true); }  // Should not be reached
+            ]
+        };
+        var agent = new ToolApprovalAgent(innerAgent.Object, options);
+
+        // Act
+        await agent.RunAsync([new ChatMessage(ChatRole.User, "Hi")], session);
+
+        // Assert — first rule matched, second was never called
+        Assert.True(rule1Called);
+        Assert.False(rule2Called);
+    }
+
+    /// <summary>
+    /// Verify that standing rules are evaluated before auto-approval rules.
+    /// </summary>
+    [Fact]
+    public async Task RunAsync_StandingRuleTakesPrecedenceOverAutoApprovalRuleAsync()
+    {
+        // Arrange
+        var session = new ChatClientAgentSession();
+        var approvalRequest = new ToolApprovalRequestContent("req1", new FunctionCallContent("call1", "MyTool"));
+
+        var callCount = 0;
+        var innerAgent = new Mock<AIAgent>();
+        innerAgent
+            .Protected()
+            .Setup<Task<AgentResponse>>("RunCoreAsync",
+                ItExpr.IsAny<IEnumerable<ChatMessage>>(),
+                ItExpr.IsAny<AgentSession?>(),
+                ItExpr.IsAny<AgentRunOptions?>(),
+                ItExpr.IsAny<CancellationToken>())
+            .ReturnsAsync(() =>
+            {
+                callCount++;
+                if (callCount <= 2)
+                {
+                    return new AgentResponse([new ChatMessage(ChatRole.Assistant, [approvalRequest])]);
+                }
+
+                return new AgentResponse([new ChatMessage(ChatRole.Assistant, "Done")]);
+            });
+
+        var heuristicCalled = false;
+        var options = new ToolApprovalAgentOptions
+        {
+            AutoApprovalRules = [fcc => { heuristicCalled = true; return new ValueTask<bool>(true); }]
+        };
+        var agent = new ToolApprovalAgent(innerAgent.Object, options);
+
+        // Call 1: heuristic should be called (no standing rule yet)
+        var response1 = await agent.RunAsync([new ChatMessage(ChatRole.User, "Hi")], session);
+        Assert.True(heuristicCalled);
+        Assert.Equal("Done", response1.Text);
+
+        // Now establish a standing rule by sending AlwaysApprove
+        heuristicCalled = false;
+        callCount = 0;
+        var alwaysApprove = new AlwaysApproveToolApprovalResponseContent(
+            approvalRequest.CreateResponse(approved: true),
+            alwaysApproveTool: true,
+            alwaysApproveToolWithArguments: false);
+
+        // Call 2: standing rule should match first, heuristic should NOT be called
+        var response2 = await agent.RunAsync(
+            [new ChatMessage(ChatRole.User, [alwaysApprove])],
+            session);
+        Assert.False(heuristicCalled);
+        Assert.Equal("Done", response2.Text);
+    }
+
+    /// <summary>
+    /// Verify that when a batch contains a mix of heuristic-approved and standing-rule-approved
+    /// requests, the collected approval responses preserve the original request order rather than
+    /// being grouped by approval kind.
+    /// </summary>
+    [Fact]
+    public async Task RunAsync_MixedAutoApprovals_PreserveOriginalOrderAsync()
+    {
+        // Arrange
+        var session = new ChatClientAgentSession();
+
+        // Batch ordering: first request is approved by a heuristic, second by a standing rule.
+        var heuristicRequest = new ToolApprovalRequestContent("reqA", new FunctionCallContent("callA", "HeuristicTool"));
+        var standingRequest = new ToolApprovalRequestContent("reqB", new FunctionCallContent("callB", "StandingTool"));
+
+        var batchResponse = new AgentResponse([new ChatMessage(ChatRole.Assistant, [heuristicRequest, standingRequest])]);
+        var finalResponse = new AgentResponse([new ChatMessage(ChatRole.Assistant, "Done")]);
+
+        var callCount = 0;
+        List<ChatMessage>? secondCallMessages = null;
+        var innerAgent = new Mock<AIAgent>();
+        innerAgent
+            .Protected()
+            .Setup<Task<AgentResponse>>("RunCoreAsync",
+                ItExpr.IsAny<IEnumerable<ChatMessage>>(),
+                ItExpr.IsAny<AgentSession?>(),
+                ItExpr.IsAny<AgentRunOptions?>(),
+                ItExpr.IsAny<CancellationToken>())
+            .Callback<IEnumerable<ChatMessage>, AgentSession?, AgentRunOptions?, CancellationToken>((msgs, _, _, _) =>
+            {
+                callCount++;
+                if (callCount == 2)
+                {
+                    secondCallMessages = msgs.ToList();
+                }
+            })
+            .ReturnsAsync(() => callCount == 1 ? batchResponse : finalResponse);
+
+        var options = new ToolApprovalAgentOptions
+        {
+            AutoApprovalRules = [fcc => new ValueTask<bool>(fcc.Name == "HeuristicTool")]
+        };
+        var agent = new ToolApprovalAgent(innerAgent.Object, options);
+
+        // Establish a standing rule for "StandingTool" via an AlwaysApprove response in the same call.
+        var alwaysApprove = standingRequest.CreateAlwaysApproveToolResponse("User said always");
+
+        // Act — both requests auto-approve (heuristic + standing rule), so the inner agent is re-invoked.
+        var response = await agent.RunAsync(
+            [new ChatMessage(ChatRole.User, [alwaysApprove])],
+            session);
+
+        // Assert — inner agent re-called and final response returned.
+        Assert.Equal(2, callCount);
+        Assert.Equal("Done", response.Text);
+
+        // The injected approval responses must preserve the original request order: reqA before reqB,
+        // even though reqA was approved by a heuristic and reqB by a standing rule.
+        Assert.NotNull(secondCallMessages);
+        var injected = secondCallMessages!
+            .SelectMany(m => m.Contents)
+            .OfType<ToolApprovalResponseContent>()
+            .Where(r => r.RequestId is "reqA" or "reqB")
+            .ToList();
+        Assert.Equal(2, injected.Count);
+        Assert.Equal("reqA", injected[0].RequestId);
+        Assert.Equal("reqB", injected[1].RequestId);
+        Assert.All(injected, r => Assert.True(r.Approved));
+    }
+
+    /// <summary>
+    /// Verify that auto-approval rules work in the streaming path.
+    /// </summary>
+    [Fact]
+    public async Task RunStreamingAsync_AutoApprovalRule_ApprovesMatchingToolAsync()
+    {
+        // Arrange
+        var session = new ChatClientAgentSession();
+        var approvalRequest = new ToolApprovalRequestContent("req1", new FunctionCallContent("call1", "ReadTool"));
+
+        var callCount = 0;
+        var innerAgent = new Mock<AIAgent>();
+        innerAgent
+            .Protected()
+            .Setup<IAsyncEnumerable<AgentResponseUpdate>>("RunCoreStreamingAsync",
+                ItExpr.IsAny<IEnumerable<ChatMessage>>(),
+                ItExpr.IsAny<AgentSession?>(),
+                ItExpr.IsAny<AgentRunOptions?>(),
+                ItExpr.IsAny<CancellationToken>())
+            .Returns(() =>
+            {
+                callCount++;
+                if (callCount == 1)
+                {
+                    return ToAsyncEnumerableAsync([new AgentResponseUpdate(ChatRole.Assistant, [approvalRequest])]);
+                }
+
+                return ToAsyncEnumerableAsync([new AgentResponseUpdate(ChatRole.Assistant, "Done")]);
+            });
+
+        var options = new ToolApprovalAgentOptions
+        {
+            AutoApprovalRules = [fcc => new ValueTask<bool>(fcc.Name == "ReadTool")]
+        };
+        var agent = new ToolApprovalAgent(innerAgent.Object, options);
+
+        // Act
+        var updates = new List<AgentResponseUpdate>();
+        await foreach (var update in agent.RunStreamingAsync([new ChatMessage(ChatRole.User, "Hi")], session))
+        {
+            updates.Add(update);
+        }
+
+        // Assert — the approval request was auto-approved, inner agent streamed twice
+        Assert.Equal(2, callCount);
+        Assert.Single(updates);
+        Assert.Equal("Done", updates[0].Text);
+    }
+
+    #endregion
 }

dotnet/tests/Microsoft.Agents.AI.Workflows.Declarative.UnitTests/ObjectModel/ForeachExecutorTest.cs

@@ -142,6 +142,34 @@ public sealed class ForeachExecutorTest(ITestOutputHelper output) : WorkflowActi
             indexName: "CurrentIndex");
     }
 
+    [Fact]
+    public async Task ForeachTakeNextWithMultiFieldRecordAsync()
+    {
+        // Arrange
+        const string CurrentValueName = "CurrentValue";
+        this.SetVariableState(CurrentValueName);
+
+        TableDataValue tableValue = DataValue.TableFromRecords(
+            DataValue.RecordFromFields(
+                new KeyValuePair<string, DataValue>("name", new StringDataValue("Alice")),
+                new KeyValuePair<string, DataValue>("role", new StringDataValue("Engineer"))));
+
+        Foreach model = this.CreateModel(
+            displayName: nameof(ForeachTakeNextWithMultiFieldRecordAsync),
+            items: ValueExpression.Literal(tableValue),
+            valueName: CurrentValueName,
+            indexName: null);
+        ForeachExecutor action = new(model, this.State);
+
+        // Act
+        await this.ExecuteAsync(action, ForeachExecutor.Steps.Next(action.Id), action.TakeNextAsync);
+
+        // Assert
+        RecordValue currentValue = Assert.IsType<RecordValue>(this.State.Get(CurrentValueName), exactMatch: false);
+        Assert.Equal("Alice", currentValue.GetField("name").ToObject());
+        Assert.Equal("Engineer", currentValue.GetField("role").ToObject());
+    }
+
     [Fact]
     public async Task ForeachTakeLastAsync()
     {

python/packages/core/AGENTS.md

@@ -76,6 +76,19 @@ agent_framework/
 - **`SkillScriptRunner`** - Protocol for file-based script execution. Any callable matching `(skill, script, args) -> Any` satisfies it. Code-defined scripts do not use a runner.
 - **`SkillsProvider`** - Context provider (extends `ContextProvider`) that discovers file-based skills from `SKILL.md` files and/or accepts code-defined `Skill` instances. Follows progressive disclosure: advertise → load → read resources / run scripts.
 
+### Model Context Protocol (`_mcp.py`)
+
+- **`MCPTool`** - Base wrapper that owns the MCP `ClientSession` and exposes the remote server's tools as `FunctionTool`s.
+- **`MCPStdioTool`** / **`MCPStreamableHTTPTool`** / **`MCPWebsocketTool`** - Transport-specific subclasses.
+- **`MCPTaskOptions`** (experimental, `MCP_LONG_RUNNING_TASKS` feature, **frozen**) - Per-tool-instance options controlling the SEP-2663 long-running task lifecycle. When the server advertises a tool with `execution.taskSupport == "required"`, `MCPTool.call_tool` transparently routes through `call_tool_as_task`, which sends an augmented `tools/call`, polls `tasks/get` until terminal, and reinterprets `tasks/result` as a normal `CallToolResult`. Instances are immutable; replace via `MCPTool.task_options = MCPTaskOptions(...)`. Fields:
+  - `default_ttl: timedelta | None` — forwarded to the server as `params.task.ttl` (milliseconds). When `None`, the server's default applies.
+  - `cancel_remote_task_on_local_cancellation: bool = True` — only gates the `CancelledError` path. Abandonment paths (see below) always cancel.
+  - `max_task_wait: timedelta | None` — client-side deadline for the whole post-create lifecycle (poll + result fetch). When exceeded, raises `ToolExecutionException` and fires a best-effort `tasks/cancel`. `None` (default) means no client-side bound. Bounds sleeps, sends, AND reconnects via `asyncio.wait_for`.
+- **Permissive fallback**: servers that ignore the augmentation (return `CallToolResult` directly) or reject the unknown `task` field with `METHOD_NOT_FOUND` / `INVALID_PARAMS` fall back to the plain `session.call_tool(...)` path so legacy servers keep working. An unparseable success response (server accepted the augmented call but returned a payload that is neither `CreateTaskResult` nor `CallToolResult`) **does not** fall back — it raises `ToolExecutionException` to avoid double-executing a side-effecting tool.
+- **Submit-vs-track reconnect policy**: a dropped connection before a `task_id` is known raises `ToolExecutionException("connection lost; task state unknown")` without re-issuing the augmented `tools/call`, so a server that accepted the request but lost the response cannot be made to start the same operation twice; once a `task_id` exists, `tasks/get` / `tasks/result` reconnect once and retry against the same id (a shared `_send_with_one_reconnect` helper).
+- **Cancel-on-abandonment vs terminal failure**: any path where the remote task may still be running (max-wait exceeded, hard `McpError` in poll, malformed `tasks/get`, second connection loss in poll/fetch, reconnect failure) fires best-effort `tasks/cancel` before raising. Terminal failures (`failed`/`cancelled`/`input_required` server-side, `completed+isError`, malformed `tasks/result` after server completed) do **not** cancel — the server is already done. `_MCPTaskAbandoned` is the private marker distinguishing the two.
+- **Transient poll retry**: a slow `tasks/get` that surfaces as `McpError(code=408 REQUEST_TIMEOUT)` is retried (bounded by `max_task_wait`). All other non-connection `McpError`s during poll are treated as abandonment. `tasks/result` does not get transient retry — the server has already completed, so a slow payload fetch is anomalous.
+
 ### File Access Harness (`_harness/_file_access.py`)
 
 - **`AgentFileStore`** - Abstract async store backing the file-access harness. Implementations expose `write_file`, `read_file`, `delete_file`, `list_files`, `file_exists`, `search_files`, and `create_directory` over forward-slash relative paths.

python/packages/core/agent_framework/__init__.py

@@ -124,7 +124,7 @@ from ._harness._todo import (
     TodoSessionStore,
     TodoStore,
 )
-from ._mcp import MCPStdioTool, MCPStreamableHTTPTool, MCPWebsocketTool
+from ._mcp import MCPStdioTool, MCPStreamableHTTPTool, MCPTaskOptions, MCPWebsocketTool
 from ._middleware import (
     AgentContext,
     AgentMiddleware,
@@ -444,12 +444,13 @@ __all__ = [
     "InlineSkillResource",
     "InlineSkillScript",
     "LocalEvaluator",
-    "MCPStdioTool",
-    "MCPStreamableHTTPTool",
-    "MCPWebsocketTool",
     "MCPSkill",
     "MCPSkillResource",
     "MCPSkillsSource",
+    "MCPStdioTool",
+    "MCPStreamableHTTPTool",
+    "MCPTaskOptions",
+    "MCPWebsocketTool",
     "MemoryContextProvider",
     "MemoryFileStore",
     "MemoryIndexEntry",

python/packages/core/agent_framework/_feature_stage.py

@@ -58,6 +58,7 @@ class ExperimentalFeature(str, Enum):
     FOUNDRY_PREVIEW_TOOLS = "FOUNDRY_PREVIEW_TOOLS"
     FUNCTIONAL_WORKFLOWS = "FUNCTIONAL_WORKFLOWS"
     HARNESS = "HARNESS"
+    MCP_LONG_RUNNING_TASKS = "MCP_LONG_RUNNING_TASKS"
     MCP_SKILLS = "MCP_SKILLS"
     PROGRESSIVE_TOOLS = "PROGRESSIVE_TOOLS"
     SKILLS = "SKILLS"

python/packages/core/agent_framework/_workflows/_agent.py

@@ -2,7 +2,6 @@
 
 from __future__ import annotations
 
-import json
 import logging
 import sys
 import uuid
@@ -12,7 +11,6 @@ from datetime import datetime, timezone
 from typing import TYPE_CHECKING, Any, ClassVar, Literal, cast, overload
 
 from .._agents import BaseAgent
-from .._serialization import make_json_safe
 from .._sessions import (
     AgentSession,
     ContextProvider,
@@ -30,11 +28,12 @@ from .._types import (
     UsageDetails,
     add_usage_details,
 )
-from ..exceptions import AgentInvalidRequestException, AgentInvalidResponseException
+from ..exceptions import AgentException, AgentInvalidRequestException, AgentInvalidResponseException
 from ._checkpoint import CheckpointStorage
 from ._events import (
     AGENT_FORWARDED_EVENT_TYPES,
     WorkflowEvent,
+    WorkflowRunState,
 )
 from ._message_utils import normalize_messages_input
 from ._typing_utils import is_instance_of, is_type_compatible
@@ -59,27 +58,24 @@ class WorkflowAgent(BaseAgent):
     @dataclass
     class RequestInfoFunctionArgs:
         request_id: str
-        data: Any
+        request_event: WorkflowEvent
 
         def to_dict(self) -> dict[str, Any]:
-            return {"request_id": self.request_id, "data": make_json_safe(self.data)}
-
-        def to_json(self) -> str:
-            return json.dumps(self.to_dict())
+            return {"request_id": self.request_id, "request_event": self.request_event.to_dict()}
 
         @classmethod
         def from_dict(cls, payload: dict[str, Any]) -> WorkflowAgent.RequestInfoFunctionArgs:
-            return cls(request_id=payload.get("request_id", ""), data=payload.get("data"))
+            if "request_id" not in payload or "request_event" not in payload:
+                raise ValueError(
+                    "Invalid payload for RequestInfoFunctionArgs. 'request_id' and 'request_event' are required."
+                )
+            if not payload["request_id"]:
+                raise ValueError("request_id cannot be empty.")
 
-        @classmethod
-        def from_json(cls, raw: str) -> WorkflowAgent.RequestInfoFunctionArgs:
-            try:
-                parsed: Any = json.loads(raw)
-            except json.JSONDecodeError as exc:
-                raise ValueError(f"RequestInfoFunctionArgs JSON payload is malformed: {exc}") from exc
-            if not isinstance(parsed, dict):
-                raise ValueError("RequestInfoFunctionArgs JSON payload must decode to a mapping")
-            return cls.from_dict(cast(dict[str, Any], parsed))
+            return cls(
+                request_id=payload.get("request_id", ""),
+                request_event=WorkflowEvent.from_dict(payload.get("request_event", {})),
+            )
 
     def __init__(
         self,
@@ -129,16 +125,11 @@ class WorkflowAgent(BaseAgent):
             **kwargs,
         )
         self._workflow: Workflow = workflow
-        self._pending_requests: dict[str, WorkflowEvent[Any]] = {}
 
     @property
     def workflow(self) -> Workflow:
         return self._workflow
 
-    @property
-    def pending_requests(self) -> dict[str, WorkflowEvent[Any]]:
-        return self._pending_requests
-
     # region Run Methods
 
     @overload
@@ -182,7 +173,7 @@ class WorkflowAgent(BaseAgent):
 
         Args:
             messages: The message(s) to send to the workflow. Required for new runs,
-                should be None when resuming from checkpoint.
+                could be None if only restoring the underlying workflow from a checkpoint.
 
         Keyword Args:
             stream: If True, returns an async iterable of updates. If False (default),
@@ -416,101 +407,79 @@ class WorkflowAgent(BaseAgent):
         Yields:
             WorkflowEvent objects from the workflow execution.
         """
-        # Determine the execution mode based on state.
-        # The streaming flag controls the workflow's internal streaming mode,
-        # which affects executor behavior (e.g. AgentExecutor emits different event
-        # types in streaming vs non-streaming mode).
-        if bool(self.pending_requests):
-            function_responses = self._process_pending_requests(input_messages)
+        # Restore the workflow state if a checkpoint is provided
+        if checkpoint_id is not None:
+            if checkpoint_storage is None:
+                raise AgentInvalidRequestException("checkpoint_storage must be provided when checkpoint_id is provided")
+            logger.debug(f"Restoring workflow from checkpoint {checkpoint_id}")
+            # Restore the workflow from checkpoint
             if streaming:
-                async for event in self.workflow.run(
-                    responses=function_responses,
-                    stream=True,
-                    function_invocation_kwargs=function_invocation_kwargs,
-                    client_kwargs=client_kwargs,
-                ):
-                    yield event
-            else:
-                for event in await self.workflow.run(
-                    responses=function_responses,
-                    function_invocation_kwargs=function_invocation_kwargs,
-                    client_kwargs=client_kwargs,
-                ):
-                    yield event
-
-        elif checkpoint_id is not None:
-            # Restore the prior workflow state from the checkpoint. Shared
-            # state (e.g. accumulated conversation history maintained by the
-            # workflow's executors) survives across turns because Workflow.run
-            # no longer wipes state per call. Callers who want to deliver a
-            # new user message after restore should make a second
-            # `workflow.run(message=...)` call - they are NOT mutually
-            # exclusive on the same instance, but each must be its own call.
-            if streaming:
-                async for event in self.workflow.run(
+                async for _ in self.workflow.run(
                     stream=True,
                     checkpoint_id=checkpoint_id,
                     checkpoint_storage=checkpoint_storage,
+                ):
+                    pass
+            else:
+                _ = await self.workflow.run(
+                    checkpoint_id=checkpoint_id,
+                    checkpoint_storage=checkpoint_storage,
+                )
+            if not input_messages:
+                logger.info("No input messages provided; the workflow has been restored to the checkpoint state.")
+                return
+
+        final_state = self._workflow.status
+        logger.debug(f"Workflow state: {final_state}")
+
+        if final_state == WorkflowRunState.IDLE_WITH_PENDING_REQUESTS:
+            # Extract function responses from input messages, and ensure that
+            # only function responses are present in messages if there is any
+            # pending request.
+            # NOTE: It is possible that some pending requests are not fulfilled,
+            # and we will let the workflow to handle this -- the agent does not
+            # have an opinion on this.
+            function_responses = self._extract_function_responses(input_messages)
+            if streaming:
+                async for event in self.workflow.run(
+                    responses=function_responses,
+                    stream=True,
+                    checkpoint_storage=checkpoint_storage,
                     function_invocation_kwargs=function_invocation_kwargs,
                     client_kwargs=client_kwargs,
                 ):
                     yield event
             else:
                 for event in await self.workflow.run(
-                    checkpoint_id=checkpoint_id,
+                    responses=function_responses,
+                    checkpoint_storage=checkpoint_storage,
+                    function_invocation_kwargs=function_invocation_kwargs,
+                    client_kwargs=client_kwargs,
+                ):
+                    yield event
+        elif final_state == WorkflowRunState.IDLE:
+            if streaming:
+                async for event in self.workflow.run(
+                    message=input_messages,
+                    stream=True,
+                    checkpoint_storage=checkpoint_storage,
+                    function_invocation_kwargs=function_invocation_kwargs,
+                    client_kwargs=client_kwargs,
+                ):
+                    yield event
+            else:
+                for event in await self.workflow.run(
+                    message=input_messages,
                     checkpoint_storage=checkpoint_storage,
                     function_invocation_kwargs=function_invocation_kwargs,
                     client_kwargs=client_kwargs,
                 ):
                     yield event
-
         else:
-            if streaming:
-                async for event in self.workflow.run(
-                    message=input_messages,
-                    stream=True,
-                    checkpoint_storage=checkpoint_storage,
-                    function_invocation_kwargs=function_invocation_kwargs,
-                    client_kwargs=client_kwargs,
-                ):
-                    yield event
-            else:
-                for event in await self.workflow.run(
-                    message=input_messages,
-                    checkpoint_storage=checkpoint_storage,
-                    function_invocation_kwargs=function_invocation_kwargs,
-                    client_kwargs=client_kwargs,
-                ):
-                    yield event
+            raise AgentException(f"The underlying workflow is in an invalid state to restart: {final_state}.")
 
     # endregion Run Methods
 
-    def _process_pending_requests(self, input_messages: Sequence[Message]) -> dict[str, Any]:
-        """Process pending requests by extracting function responses and updating state.
-
-        Args:
-            input_messages: Input messages that may contain function responses.
-
-        Returns:
-            A dictionary mapping request IDs to their response data.
-        """
-        logger.info(f"Continuing workflow to address {len(self.pending_requests)} requests")
-
-        # Extract function responses from input messages, and ensure that
-        # only function responses are present in messages if there is any
-        # pending request.
-        function_responses = self._extract_function_responses(input_messages)
-
-        # Pop pending requests if fulfilled.
-        for request_id in list(self.pending_requests.keys()):
-            if request_id in function_responses:
-                self.pending_requests.pop(request_id)
-
-        # NOTE: It is possible that some pending requests are not fulfilled,
-        # and we will let the workflow to handle this -- the agent does not
-        # have an opinion on this.
-        return function_responses
-
     def _convert_workflow_events_to_agent_response(
         self,
         response_id: str,
@@ -528,10 +497,10 @@ class WorkflowAgent(BaseAgent):
 
         for output_event in output_events:
             if output_event.type == "request_info":
-                function_call, approval_request = self._process_request_info_event(output_event)
+                request_content = self._process_request_info_event(output_event)
                 messages.append(
                     Message(
-                        contents=[function_call, approval_request],
+                        contents=[request_content],
                         role="assistant",
                         author_name=output_event.source_executor_id,
                         message_id=str(uuid.uuid4()),
@@ -598,38 +567,6 @@ class WorkflowAgent(BaseAgent):
             raw_representation=raw_representations,
         )
 
-    def _process_request_info_event(
-        self,
-        event: WorkflowEvent[Any],
-    ) -> tuple[Content, Content]:
-        """Convert a request_info event to FunctionCallContent and FunctionApprovalRequestContent.
-
-        Args:
-            event: A WorkflowEvent with type='request_info'.
-
-        Returns:
-            A tuple of (FunctionCallContent, FunctionApprovalRequestContent).
-        """
-        request_id = event.request_id
-        if not request_id:
-            raise ValueError("request_info event must have a request_id")
-
-        self.pending_requests[request_id] = event
-
-        args = self.RequestInfoFunctionArgs(request_id=request_id, data=event.data).to_dict()
-
-        function_call = Content.from_function_call(
-            call_id=request_id,
-            name=self.REQUEST_INFO_FUNCTION_NAME,
-            arguments=args,
-        )
-        approval_request = Content.from_function_approval_request(
-            id=request_id,
-            function_call=function_call,
-            additional_properties={"request_id": request_id},
-        )
-        return function_call, approval_request
-
     def _convert_workflow_event_to_agent_response_updates(
         self,
         response_id: str,
@@ -731,85 +668,72 @@ class WorkflowAgent(BaseAgent):
             ]
 
         if event.type == "request_info":
-            # Store the pending request for later correlation
-            request_id = event.request_id
-            if not request_id:
-                raise ValueError("request_info event must have a request_id")
-
-            self.pending_requests[request_id] = event
-
-            args = self.RequestInfoFunctionArgs(request_id=request_id, data=event.data).to_dict()
-
-            function_call = Content.from_function_call(
-                call_id=request_id,
-                name=self.REQUEST_INFO_FUNCTION_NAME,
-                arguments=args,
-            )
-            approval_request = Content.from_function_approval_request(
-                id=request_id,
-                function_call=function_call,
-                additional_properties={"request_id": request_id},
-            )
+            request_content = self._process_request_info_event(event)
             return [
                 AgentResponseUpdate(
-                    contents=[function_call, approval_request],
+                    contents=[request_content],
                     role="assistant",
                     author_name=self.name,
                     response_id=response_id,
                     message_id=str(uuid.uuid4()),
                     created_at=datetime.now(tz=timezone.utc).strftime("%Y-%m-%dT%H:%M:%S.%fZ"),
+                    raw_representation=event,
                 )
             ]
 
         # Ignore workflow-internal events
         return []
 
+    def _process_request_info_event(
+        self,
+        event: WorkflowEvent[Any],
+    ) -> Content:
+        """Convert a request_info event to FunctionApprovalRequestContent.
+
+        Args:
+            event: A WorkflowEvent with type='request_info'.
+
+        Returns:
+            A content object representing the request info. The content can be a `function_approval_request`
+            or a `function_call` depending on the structure of the event data.
+
+        Note:
+            If the event data is already a FunctionApprovalRequestContent, it will be returned as-is.
+        """
+        if isinstance(event.data, Content) and event.data.user_input_request:
+            # Return the event data as-is if it's already a properly formed FunctionApprovalRequestContent
+            return event.data
+
+        request_id = event.request_id
+        args = self.RequestInfoFunctionArgs(request_id=request_id, request_event=event).to_dict()
+
+        return Content.from_function_call(
+            call_id=request_id,
+            name=self.REQUEST_INFO_FUNCTION_NAME,
+            arguments=args,
+        )
+
     def _extract_function_responses(self, input_messages: Sequence[Message]) -> dict[str, Any]:
-        """Extract function responses from input messages."""
+        """Extract function responses from input messages.
+
+        The responses are for pending requests that the workflow is waiting on, and
+        will be passed to the workflow. The pending requests are processed to either
+        `function_approval_request` or `function_call` content by `_process_request_info_event`.
+        """
         function_responses: dict[str, Any] = {}
         for message in input_messages:
             for content in message.contents:
                 if content.type == "function_approval_response":
-                    # Parse the function arguments to recover request payload
-                    arguments_payload = content.function_call.arguments  # type: ignore[attr-defined, union-attr]
-                    if isinstance(arguments_payload, str):
-                        try:
-                            parsed_args = self.RequestInfoFunctionArgs.from_json(arguments_payload)
-                        except ValueError as exc:
-                            raise AgentInvalidResponseException(
-                                "FunctionApprovalResponseContent arguments must decode to a mapping."
-                            ) from exc
-                    elif isinstance(arguments_payload, dict):
-                        parsed_args = self.RequestInfoFunctionArgs.from_dict(arguments_payload)
-                    else:
-                        raise AgentInvalidResponseException(
-                            "FunctionApprovalResponseContent arguments must be a mapping or JSON string."
-                        )
-
-                    request_id = parsed_args.request_id or content.id  # type: ignore[attr-defined]
-                    if not content.approved:  # type: ignore[attr-defined]
-                        raise AgentInvalidResponseException(f"Request '{request_id}' was not approved by the caller.")
-
-                    if request_id in self.pending_requests:
-                        function_responses[request_id] = parsed_args.data
-                    elif bool(self.pending_requests):
-                        raise AgentInvalidRequestException(
-                            "Only responses for pending requests are allowed when there are outstanding approvals."
-                        )
+                    request_id: str = content.id  # type: ignore[assignment]
+                    function_responses[request_id] = content
                 elif content.type == "function_result":
-                    request_id = content.call_id  # type: ignore[attr-defined]
-                    if request_id in self.pending_requests:
-                        response_data = content.result if hasattr(content, "result") else str(content)  # type: ignore[attr-defined]
-                        function_responses[request_id] = response_data
-                    elif bool(self.pending_requests):
-                        raise AgentInvalidRequestException(
-                            "Only function responses for pending requests are allowed while requests are outstanding."
-                        )
+                    response_data = content.result if hasattr(content, "result") else str(content)  # type: ignore[attr-defined]
+                    function_responses[content.call_id] = response_data  # type: ignore
                 else:
-                    if bool(self.pending_requests):
-                        raise AgentInvalidResponseException(
-                            "Unexpected content type while awaiting request info responses."
-                        )
+                    raise AgentInvalidResponseException(
+                        "Unexpected content type while awaiting request info responses."
+                    )
+
         return function_responses
 
     def _extract_contents(self, data: Any) -> list[Content]:

python/packages/core/agent_framework/_workflows/_agent_executor.py

@@ -429,15 +429,30 @@ class AgentExecutor(Executor):
             function_invocation_kwargs=function_invocation_kwargs,
             client_kwargs=client_kwargs,
         )
-        await ctx.yield_output(response)
 
         # Handle any user input requests
         if response.user_input_requests:
+            user_input_request_count = len(response.user_input_requests)
+            total_message_content_count = sum(len(msg.contents) for msg in response.messages)
+            if user_input_request_count != total_message_content_count:
+                logger.warning(
+                    "Response %s contains %d user input requests but total message contents are %d. "
+                    "This indicates the response contains both user input requests and message contents. "
+                    "Double check if this is the intended behavior, as non user input request contents in "
+                    "this response will not be emitted.",
+                    response.response_id,
+                    user_input_request_count,
+                    total_message_content_count,
+                )
             for user_input_request in response.user_input_requests:
                 self._pending_agent_requests[user_input_request.id] = user_input_request  # type: ignore[index]
-                await ctx.request_info(user_input_request, Content)
+                await ctx.request_info(user_input_request, Content, request_id=user_input_request.id)
             return None
 
+        # Only yield output if the response is complete and not waiting for user input.
+        # This is to avoid emitting two events of different types ('output' and 'request_info')
+        # that carry the same payload.
+        await ctx.yield_output(response)
         return response
 
     async def _run_agent_streaming(self, ctx: WorkflowContext[Never, AgentResponseUpdate]) -> AgentResponse | None:
@@ -472,9 +487,25 @@ class AgentExecutor(Executor):
         )
         async for update in stream:
             updates.append(update)
-            await ctx.yield_output(update)
             if update.user_input_requests:
+                user_input_request_count = len(update.user_input_requests)
+                total_message_content_count = len(update.contents)
+                if user_input_request_count != total_message_content_count:
+                    logger.warning(
+                        "Response update %s contains %d user input requests but total message contents are %d. "
+                        "This indicates the response update contains both user input requests and message contents. "
+                        "Double check if this is the intended behavior, as non user input request contents will "
+                        "not be emitted.",
+                        update.response_id,
+                        user_input_request_count,
+                        total_message_content_count,
+                    )
                 streamed_user_input_requests.extend(update.user_input_requests)
+            else:
+                # Only yield output events for updates that do not contain user input requests.
+                # This is to avoid emitting two events of different types ('output' and 'request_info')
+                # that carry the same payload.
+                await ctx.yield_output(update)
 
         # Prefer stream finalization when available so result hooks run
         # (e.g., thread conversation updates). Fall back to reconstructing from updates
@@ -509,7 +540,7 @@ class AgentExecutor(Executor):
         if user_input_requests:
             for user_input_request in user_input_requests:
                 self._pending_agent_requests[user_input_request.id] = user_input_request  # type: ignore[index]
-                await ctx.request_info(user_input_request, Content)
+                await ctx.request_info(user_input_request, Content, request_id=user_input_request.id)
             return None
 
         return response

python/packages/core/agent_framework/_workflows/_workflow.py

@@ -360,6 +360,22 @@ class Workflow(DictConvertible):
         # Flag to prevent concurrent workflow executions
         self._is_running = False
 
+        # Current run-level status of this workflow instance. Updated in lockstep with
+        # the status events emitted from `_run_workflow_with_tracing`. Defaults to IDLE
+        # for a freshly built workflow that has not yet been run.
+        self._status: WorkflowRunState = WorkflowRunState.IDLE
+
+    @property
+    def status(self) -> WorkflowRunState:
+        """Return the current run-level status of this workflow instance.
+
+        Mirrors the most recent status event emitted by the workflow. Safe to read at
+        any time: workflows run on a single asyncio event loop, and the underlying
+        attribute is a single enum reference whose assignment is atomic under the
+        CPython GIL, so no locking is required.
+        """
+        return self._status
+
     def _ensure_not_running(self) -> None:
         """Ensure the workflow is not already running."""
         if self._is_running:
@@ -513,8 +529,9 @@ class Workflow(DictConvertible):
                 with _framework_event_origin():
                     started = WorkflowEvent.started()
                 yield started  # noqa: RUF070
+                self._status = WorkflowRunState.IN_PROGRESS
                 with _framework_event_origin():
-                    in_progress = WorkflowEvent.status(WorkflowRunState.IN_PROGRESS)
+                    in_progress = WorkflowEvent.status(self._status)
                 yield in_progress  # noqa: RUF070
 
                 # Per-run reset for fresh-message runs only. We deliberately
@@ -569,17 +586,20 @@ class Workflow(DictConvertible):
 
                     if event.type == "request_info" and not emitted_in_progress_pending:
                         emitted_in_progress_pending = True
+                        self._status = WorkflowRunState.IN_PROGRESS_PENDING_REQUESTS
                         with _framework_event_origin():
-                            pending_status = WorkflowEvent.status(WorkflowRunState.IN_PROGRESS_PENDING_REQUESTS)
+                            pending_status = WorkflowEvent.status(self._status)
                         yield pending_status  # noqa: RUF070
                 # Workflow runs until idle - emit final status based on whether requests are pending
                 if saw_request:
+                    self._status = WorkflowRunState.IDLE_WITH_PENDING_REQUESTS
                     with _framework_event_origin():
-                        terminal_status = WorkflowEvent.status(WorkflowRunState.IDLE_WITH_PENDING_REQUESTS)
+                        terminal_status = WorkflowEvent.status(self._status)
                     yield terminal_status
                 else:
+                    self._status = WorkflowRunState.IDLE
                     with _framework_event_origin():
-                        terminal_status = WorkflowEvent.status(WorkflowRunState.IDLE)
+                        terminal_status = WorkflowEvent.status(self._status)
                     yield terminal_status
 
                 span.add_event(OtelAttr.WORKFLOW_COMPLETED)
@@ -593,6 +613,7 @@ class Workflow(DictConvertible):
                 with _framework_event_origin():
                     failed_event = WorkflowEvent.failed(details)
                 yield failed_event  # noqa: RUF070
+                self._status = WorkflowRunState.FAILED
                 with _framework_event_origin():
                     failed_status = WorkflowEvent.status(WorkflowRunState.FAILED)
                 yield failed_status  # noqa: RUF070

python/packages/core/agent_framework/observability.py

@@ -80,6 +80,7 @@ __all__ = [
     "EmbeddingTelemetryLayer",
     "OtelAttr",
     "configure_otel_providers",
+    "create_mcp_client_span",
     "create_metric_views",
     "create_resource",
     "disable_instrumentation",
@@ -87,6 +88,7 @@ __all__ = [
     "enable_sensitive_telemetry",
     "get_meter",
     "get_tracer",
+    "set_mcp_span_error",
 ]
 
 
@@ -110,7 +112,6 @@ INNER_ACCUMULATED_USAGE: Final[contextvars.ContextVar[UsageDetails | None]] = co
     "inner_accumulated_usage", default=None
 )
 
-
 OTEL_METRICS: Final[str] = "__otel_metrics__"
 TOKEN_USAGE_BUCKET_BOUNDARIES: Final[tuple[float, ...]] = (
     1,
@@ -292,6 +293,14 @@ class OtelAttr(str, Enum):
     AGENT_CREATE_OPERATION = "create_agent"
     AGENT_INVOKE_OPERATION = "invoke_agent"
 
+    # MCP attributes (https://opentelemetry.io/docs/specs/semconv/gen-ai/mcp/)
+    MCP_METHOD_NAME = "mcp.method.name"
+    MCP_PROTOCOL_VERSION = "mcp.protocol.version"
+    MCP_SESSION_ID = "mcp.session.id"
+    PROMPT_NAME = "gen_ai.prompt.name"
+    NETWORK_TRANSPORT = "network.transport"
+    NETWORK_PROTOCOL_NAME = "network.protocol.name"
+
     # Agent Framework specific attributes
     MEASUREMENT_FUNCTION_TAG_NAME = "agent_framework.function.name"
     MEASUREMENT_FUNCTION_INVOCATION_DURATION = "agent_framework.function.invocation.duration"
@@ -2013,6 +2022,61 @@ def get_function_span(
     )
 
 
+# region MCP span helpers
+
+
+@contextlib.contextmanager
+def create_mcp_client_span(
+    method_name: str,
+    target: str | None = None,
+    attributes: dict[str, Any] | None = None,
+) -> Generator[trace.Span, Any, Any]:
+    """Create an MCP client span per OTel MCP semantic conventions.
+
+    Span name follows the format ``{mcp.method.name} {target}`` when a target
+    is available, otherwise just ``{mcp.method.name}``.
+
+    See: https://opentelemetry.io/docs/specs/semconv/gen-ai/mcp/#client
+
+    Args:
+        method_name: The MCP method name (e.g. ``initialize``, ``tools/call``).
+        target: Optional low-cardinality target (tool name, prompt name).
+        attributes: Additional span attributes.
+    """
+    span_name = f"{method_name} {target}" if target else method_name
+    attrs: dict[str, Any] = {OtelAttr.MCP_METHOD_NAME: method_name}
+    if attributes:
+        attrs.update(attributes)
+    tracer = get_tracer() if OBSERVABILITY_SETTINGS.ENABLED else trace.NoOpTracer()
+    span = tracer.start_span(span_name, kind=trace.SpanKind.CLIENT, attributes=attrs)
+    with trace.use_span(
+        span=span,
+        end_on_exit=True,
+        record_exception=True,
+        set_status_on_exception=True,
+    ) as current_span:
+        yield current_span
+
+
+def set_mcp_span_error(
+    span: trace.Span,
+    error_type: str,
+    description: str | None = None,
+) -> None:
+    """Set error status and ``error.type`` on an MCP span.
+
+    Args:
+        span: The span to mark as errored.
+        error_type: The error type string (e.g. ``tool_error``, exception class name).
+        description: Optional description (e.g. JSON-RPC error message).
+    """
+    span.set_attribute(OtelAttr.ERROR_TYPE, error_type)
+    span.set_status(trace.StatusCode.ERROR, description=description)
+
+
+# endregion
+
+
 @contextlib.contextmanager
 def _activate_span(span: trace.Span) -> Generator[None]:
     """Attach ``span`` as the current span in the OpenTelemetry context.

python/packages/core/tests/core/test_mcp_observability.py

@@ -0,0 +1,376 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+"""Tests for MCP client span instrumentation per OTel GenAI Semantic Conventions.
+
+See: https://opentelemetry.io/docs/specs/semconv/gen-ai/mcp/#client
+"""
+
+from __future__ import annotations
+
+from typing import Any
+from unittest.mock import AsyncMock, Mock
+
+import pytest
+from mcp import types
+from mcp.shared.exceptions import McpError
+from mcp.types import ErrorData
+from opentelemetry.sdk.trace.export.in_memory_span_exporter import InMemorySpanExporter
+from opentelemetry.trace import SpanKind, StatusCode
+
+from agent_framework import MCPStdioTool, MCPStreamableHTTPTool, MCPWebsocketTool
+from agent_framework._mcp import MCPTool
+from agent_framework.exceptions import ToolExecutionException
+from agent_framework.observability import OtelAttr
+
+# region helpers
+
+
+def _make_connected_mcp_tool(
+    name: str = "test-mcp",
+    *,
+    supports_tools: bool = True,
+    supports_prompts: bool = True,
+) -> MCPTool:
+    """Create an MCPTool with a mocked session, ready for testing."""
+    tool = MCPTool(name=name)
+    tool.session = AsyncMock()
+    tool.is_connected = True
+    tool._supports_tools = supports_tools
+    tool._supports_prompts = supports_prompts
+    tool.load_tools_flag = True
+    tool.load_prompts_flag = True
+    return tool
+
+
+def _make_tool_list_result(
+    tools: list[dict[str, Any]] | None = None,
+) -> Mock:
+    """Create a mock ListToolsResult."""
+    if tools is None:
+        tools = [{"name": "get-weather", "description": "Get weather", "inputSchema": {"type": "object"}}]
+    result = Mock()
+    result.tools = [
+        types.Tool(name=t["name"], description=t.get("description", ""), inputSchema=t.get("inputSchema", {}))
+        for t in tools
+    ]
+    result.nextCursor = None
+    return result
+
+
+def _make_prompt_list_result(
+    prompts: list[dict[str, Any]] | None = None,
+) -> Mock:
+    """Create a mock ListPromptsResult."""
+    if prompts is None:
+        prompts = [{"name": "analyze-code", "description": "Analyze code"}]
+    result = Mock()
+    result.prompts = [
+        types.Prompt(name=p["name"], description=p.get("description", ""), arguments=None) for p in prompts
+    ]
+    result.nextCursor = None
+    return result
+
+
+def _make_call_tool_result(text: str = "result", is_error: bool = False) -> Mock:
+    """Create a mock CallToolResult."""
+    result = Mock()
+    result.isError = is_error
+    result.content = [types.TextContent(type="text", text=text)]
+    return result
+
+
+def _make_get_prompt_result(text: str = "prompt result") -> types.GetPromptResult:
+    """Create a mock GetPromptResult."""
+    return types.GetPromptResult(
+        description="test prompt",
+        messages=[
+            types.PromptMessage(
+                role="user",
+                content=types.TextContent(type="text", text=text),
+            )
+        ],
+    )
+
+
+# endregion
+
+
+# region initialize span
+
+
+async def test_mcp_initialize_span(span_exporter: InMemorySpanExporter):
+    """session.initialize() should produce an MCP CLIENT span named 'initialize'."""
+    tool = MCPTool(name="test-server")
+
+    mock_session_cls = AsyncMock()
+    init_result = Mock()
+    init_result.capabilities = None
+    init_result.protocolVersion = "2025-06-18"
+    mock_session_cls.initialize = AsyncMock(return_value=init_result)
+
+    # Create a mock transport context manager
+    mock_transport = AsyncMock()
+    mock_transport.__aenter__ = AsyncMock(return_value=(Mock(), Mock()))
+    mock_transport.__aexit__ = AsyncMock(return_value=False)
+
+    # Mock get_mcp_client and the session creation
+    tool.session = None
+    tool.load_tools_flag = False
+    tool.load_prompts_flag = False
+
+    span_exporter.clear()
+
+    with pytest.MonkeyPatch.context() as m:
+        m.setattr(tool, "get_mcp_client", lambda: mock_transport)
+
+        async def patched_connect(self_: Any, *, reset: bool = False, load_configured: bool = True) -> None:
+            # Simulate _connect_on_owner: create initialize span and call session.initialize()
+            from agent_framework._mcp import create_mcp_client_span
+            from agent_framework.observability import OtelAttr
+
+            with create_mcp_client_span("initialize", attributes=self_._mcp_base_span_attributes()) as init_span:
+                result = await mock_session_cls.initialize()
+                protocol_version = getattr(result, "protocolVersion", None)
+                if protocol_version:
+                    init_span.set_attribute(OtelAttr.MCP_PROTOCOL_VERSION, protocol_version)
+
+            self_.session = mock_session_cls
+            self_.is_connected = True
+
+        m.setattr(MCPTool, "_connect_on_owner", patched_connect)
+        await tool.connect()
+
+    mock_session_cls.initialize.assert_awaited_once()
+    spans = span_exporter.get_finished_spans()
+    init_spans = [s for s in spans if s.name == "initialize"]
+    assert len(init_spans) == 1
+    span = init_spans[0]
+    assert span.kind == SpanKind.CLIENT
+    assert span.attributes[OtelAttr.MCP_METHOD_NAME] == "initialize"
+    assert span.attributes.get(OtelAttr.MCP_PROTOCOL_VERSION) == "2025-06-18"
+
+
+# endregion
+
+
+# region tools/list span
+
+
+async def test_mcp_tools_list_span(span_exporter: InMemorySpanExporter):
+    """session.list_tools() should produce an MCP CLIENT span named 'tools/list'."""
+    tool = _make_connected_mcp_tool()
+    tool.session.list_tools = AsyncMock(return_value=_make_tool_list_result())
+
+    span_exporter.clear()
+    await tool.load_tools()
+
+    spans = span_exporter.get_finished_spans()
+    list_spans = [s for s in spans if s.name == "tools/list"]
+    assert len(list_spans) == 1
+    span = list_spans[0]
+    assert span.kind == SpanKind.CLIENT
+    assert span.attributes[OtelAttr.MCP_METHOD_NAME] == "tools/list"
+
+
+# endregion
+
+
+# region prompts/list span
+
+
+async def test_mcp_prompts_list_span(span_exporter: InMemorySpanExporter):
+    """session.list_prompts() should produce an MCP CLIENT span named 'prompts/list'."""
+    tool = _make_connected_mcp_tool()
+    tool.session.list_prompts = AsyncMock(return_value=_make_prompt_list_result())
+
+    span_exporter.clear()
+    await tool.load_prompts()
+
+    spans = span_exporter.get_finished_spans()
+    list_spans = [s for s in spans if s.name == "prompts/list"]
+    assert len(list_spans) == 1
+    span = list_spans[0]
+    assert span.kind == SpanKind.CLIENT
+    assert span.attributes[OtelAttr.MCP_METHOD_NAME] == "prompts/list"
+
+
+# endregion
+
+
+# region tools/call span
+
+
+async def test_mcp_tools_call_creates_client_span_when_no_parent(span_exporter: InMemorySpanExporter):
+    """Direct call_tool() without FunctionTool wrapper creates new MCP CLIENT span."""
+    tool = _make_connected_mcp_tool()
+    tool.session.call_tool = AsyncMock(return_value=_make_call_tool_result("hello"))
+
+    span_exporter.clear()
+    result = await tool.call_tool("get-weather", city="Seattle")
+
+    assert result is not None
+    spans = span_exporter.get_finished_spans()
+    call_spans = [s for s in spans if "tools/call" in s.name]
+    assert len(call_spans) == 1
+    span = call_spans[0]
+    assert span.kind == SpanKind.CLIENT
+    assert span.name == "tools/call get-weather"
+    assert span.attributes[OtelAttr.MCP_METHOD_NAME] == "tools/call"
+    assert span.attributes[OtelAttr.TOOL_NAME] == "get-weather"
+
+
+async def test_mcp_tools_call_tool_error_sets_error_type(span_exporter: InMemorySpanExporter):
+    """When CallToolResult.isError is true, error.type should be 'tool_error' per MCP spec."""
+    tool = _make_connected_mcp_tool()
+    tool.session.call_tool = AsyncMock(return_value=_make_call_tool_result("bad input", is_error=True))
+
+    span_exporter.clear()
+    with pytest.raises(ToolExecutionException):
+        await tool.call_tool("get-weather", city="invalid")
+
+    spans = span_exporter.get_finished_spans()
+    call_spans = [s for s in spans if "tools/call" in s.name]
+    assert len(call_spans) == 1
+    span = call_spans[0]
+    assert span.attributes.get(OtelAttr.ERROR_TYPE) == "tool_error"
+    assert span.status.status_code == StatusCode.ERROR
+
+
+async def test_mcp_tools_call_mcp_error_sets_error_type(span_exporter: InMemorySpanExporter):
+    """When session.call_tool() raises McpError, error.type should be the exception class name."""
+    tool = _make_connected_mcp_tool()
+    tool.session.call_tool = AsyncMock(side_effect=McpError(ErrorData(code=-32600, message="invalid request")))
+
+    span_exporter.clear()
+    with pytest.raises(ToolExecutionException):
+        await tool.call_tool("get-weather")
+
+    spans = span_exporter.get_finished_spans()
+    call_spans = [s for s in spans if "tools/call" in s.name]
+    assert len(call_spans) == 1
+    span = call_spans[0]
+    assert span.attributes.get(OtelAttr.ERROR_TYPE) == "McpError"
+    assert span.status.status_code == StatusCode.ERROR
+
+
+# endregion
+
+
+# region prompts/get span
+
+
+async def test_mcp_prompts_get_creates_client_span(span_exporter: InMemorySpanExporter):
+    """get_prompt() should always create a new MCP CLIENT span (not enrich execute_tool)."""
+    tool = _make_connected_mcp_tool()
+    tool.session.get_prompt = AsyncMock(return_value=_make_get_prompt_result("code analysis"))
+
+    span_exporter.clear()
+    result = await tool.get_prompt("analyze-code", language="python")
+
+    assert "code analysis" in result
+    spans = span_exporter.get_finished_spans()
+    prompt_spans = [s for s in spans if "prompts/get" in s.name]
+    assert len(prompt_spans) == 1
+    span = prompt_spans[0]
+    assert span.kind == SpanKind.CLIENT
+    assert span.name == "prompts/get analyze-code"
+    assert span.attributes[OtelAttr.MCP_METHOD_NAME] == "prompts/get"
+    assert span.attributes[OtelAttr.PROMPT_NAME] == "analyze-code"
+
+
+async def test_mcp_prompts_get_mcp_error_sets_error_type(span_exporter: InMemorySpanExporter):
+    """When session.get_prompt() raises McpError, the span should have error.type and ERROR status."""
+    tool = _make_connected_mcp_tool()
+    tool.session.get_prompt = AsyncMock(
+        side_effect=McpError(ErrorData(code=-32602, message="prompt not found"))
+    )
+
+    span_exporter.clear()
+    with pytest.raises(ToolExecutionException):
+        await tool.get_prompt("missing-prompt")
+
+    spans = span_exporter.get_finished_spans()
+    prompt_spans = [s for s in spans if "prompts/get" in s.name]
+    assert len(prompt_spans) == 1
+    span = prompt_spans[0]
+    assert span.attributes.get(OtelAttr.ERROR_TYPE) == "McpError"
+    assert span.status.status_code == StatusCode.ERROR
+
+
+# endregion
+
+
+# region transport attributes
+
+
+def test_mcp_stdio_tool_transport_attributes():
+    """MCPStdioTool should have network.transport='pipe'."""
+    tool = MCPStdioTool(name="test", command="python")
+    attrs = tool._mcp_base_span_attributes()
+    assert attrs[OtelAttr.NETWORK_TRANSPORT] == "pipe"
+    assert OtelAttr.ADDRESS not in attrs
+
+
+def test_mcp_http_tool_transport_attributes():
+    """MCPStreamableHTTPTool should have tcp transport and URL-based server address/port."""
+    tool = MCPStreamableHTTPTool(name="test", url="https://api.example.com:8443/mcp")
+    attrs = tool._mcp_base_span_attributes()
+    assert attrs[OtelAttr.NETWORK_TRANSPORT] == "tcp"
+    assert attrs[OtelAttr.NETWORK_PROTOCOL_NAME] == "http"
+    assert attrs[OtelAttr.ADDRESS] == "api.example.com"
+    assert attrs[OtelAttr.PORT] == 8443
+
+
+def test_mcp_http_tool_default_port():
+    """MCPStreamableHTTPTool should default to 443 for https."""
+    tool = MCPStreamableHTTPTool(name="test", url="https://api.example.com/mcp")
+    attrs = tool._mcp_base_span_attributes()
+    assert attrs[OtelAttr.PORT] == 443
+
+
+def test_mcp_http_tool_http_default_port():
+    """MCPStreamableHTTPTool should default to 80 for http."""
+    tool = MCPStreamableHTTPTool(name="test", url="http://localhost/mcp")
+    attrs = tool._mcp_base_span_attributes()
+    assert attrs[OtelAttr.PORT] == 80
+
+
+def test_mcp_websocket_tool_transport_attributes():
+    """MCPWebsocketTool should have tcp transport and URL-based server address/port."""
+    tool = MCPWebsocketTool(name="test", url="wss://ws.example.com:9090/mcp")
+    attrs = tool._mcp_base_span_attributes()
+    assert attrs[OtelAttr.NETWORK_TRANSPORT] == "tcp"
+    assert attrs[OtelAttr.NETWORK_PROTOCOL_NAME] == "websocket"
+    assert attrs[OtelAttr.ADDRESS] == "ws.example.com"
+    assert attrs[OtelAttr.PORT] == 9090
+
+
+def test_mcp_websocket_tool_default_port():
+    """MCPWebsocketTool should default to 443 for wss."""
+    tool = MCPWebsocketTool(name="test", url="wss://ws.example.com/mcp")
+    attrs = tool._mcp_base_span_attributes()
+    assert attrs[OtelAttr.PORT] == 443
+
+
+# endregion
+
+
+# region observability disabled
+
+
+@pytest.mark.parametrize("enable_instrumentation", [False], indirect=True)
+async def test_mcp_spans_not_created_when_observability_disabled(span_exporter: InMemorySpanExporter):
+    """No MCP spans should be created when observability is disabled."""
+    tool = _make_connected_mcp_tool()
+    tool.session.list_tools = AsyncMock(return_value=_make_tool_list_result())
+    tool.session.call_tool = AsyncMock(return_value=_make_call_tool_result("ok"))
+
+    span_exporter.clear()
+    await tool.load_tools()
+    await tool.call_tool("get-weather", city="Seattle")
+
+    spans = span_exporter.get_finished_spans()
+    assert len(spans) == 0
+
+
+# endregion

python/packages/core/tests/core/test_observability.py

@@ -25,6 +25,7 @@ from agent_framework import (
     prepend_agent_framework_to_user_agent,
     tool,
 )
+from agent_framework._serialization import make_json_safe
 from agent_framework.observability import (
     ROLE_EVENT_MAP,
     AgentTelemetryLayer,
@@ -3195,17 +3196,15 @@ def test_capture_messages_with_prepared_request_info_function_call_arguments(spa
 
     from opentelemetry import trace
 
-    from agent_framework import WorkflowAgent
-
     @dataclasses.dataclass
     class HandoffRequest:
         target_agent: str
         reason: str
 
-    arguments = WorkflowAgent.RequestInfoFunctionArgs(
-        request_id="call_dc",
-        data=HandoffRequest(target_agent="helper", reason="overflow"),
-    ).to_dict()
+    arguments = {
+        "request_id": "call_dc",
+        "data": make_json_safe(HandoffRequest(target_agent="helper", reason="overflow")),
+    }
     msg = Message(
         role="assistant",
         contents=[

python/packages/core/tests/workflow/test_agent_executor.py

@@ -699,3 +699,171 @@ async def test_resolve_executor_kwargs_empty_per_executor_does_not_fallback_to_g
     resolved = {"exec_a": {}, GLOBAL_KWARGS_KEY: {"global_key": "global_val"}}
     result = executor._resolve_executor_kwargs(resolved)  # pyright: ignore[reportPrivateUsage]
     assert result == {}
+
+
+# region Tool approval emission
+
+
+class _ApprovalEmittingAgent(BaseAgent):
+    """Agent that returns a single ``function_approval_request`` Content.
+
+    Used to verify that ``AgentExecutor`` does *not* surface the approval
+    payload via both an ``output`` event and a ``request_info`` event in the
+    same superstep — only the ``request_info`` event must carry it.
+    """
+
+    def __init__(
+        self,
+        *,
+        approval_request_id: str = "apr_1",
+        tool_name: str = "delete_file",
+        tool_arguments: dict[str, Any] | None = None,
+        **kwargs: Any,
+    ):
+        super().__init__(**kwargs)
+        self._approval_request_id = approval_request_id
+        self._tool_name = tool_name
+        self._tool_arguments: dict[str, Any] = tool_arguments or {"path": "/tmp/secret.txt"}
+        self.run_count = 0
+
+    def _build_approval_content(self) -> Content:
+        function_call = Content.from_function_call(
+            call_id=self._approval_request_id,
+            name=self._tool_name,
+            arguments=self._tool_arguments,
+        )
+        return Content.from_function_approval_request(id=self._approval_request_id, function_call=function_call)
+
+    @overload
+    def run(
+        self,
+        messages: AgentRunInputs | None = ...,
+        *,
+        stream: Literal[False] = ...,
+        session: AgentSession | None = ...,
+        **kwargs: Any,
+    ) -> Awaitable[AgentResponse[Any]]: ...
+
+    @overload
+    def run(
+        self,
+        messages: AgentRunInputs | None = ...,
+        *,
+        stream: Literal[True],
+        session: AgentSession | None = ...,
+        **kwargs: Any,
+    ) -> ResponseStream[AgentResponseUpdate, AgentResponse[Any]]: ...
+
+    def run(
+        self,
+        messages: AgentRunInputs | None = None,
+        *,
+        stream: bool = False,
+        session: AgentSession | None = None,
+        **kwargs: Any,
+    ) -> Awaitable[AgentResponse[Any]] | ResponseStream[AgentResponseUpdate, AgentResponse[Any]]:
+        self.run_count += 1
+        approval = self._build_approval_content()
+
+        if stream:
+
+            async def _stream() -> AsyncIterable[AgentResponseUpdate]:
+                yield AgentResponseUpdate(contents=[approval], role="assistant")
+
+            return ResponseStream(_stream(), finalizer=AgentResponse.from_updates)
+
+        async def _run() -> AgentResponse:
+            return AgentResponse(messages=[Message("assistant", [approval])])
+
+        return _run()
+
+
+def _has_approval_payload(event: WorkflowEvent[Any]) -> bool:
+    """Return True if the event's data carries a ``function_approval_request`` content."""
+    data: Any = event.data
+
+    def _contents_of(value: Any) -> list[Content]:
+        if isinstance(value, AgentResponseUpdate):
+            return list(value.contents)
+        if isinstance(value, AgentResponse):
+            return [c for m in value.messages for c in m.contents]
+        if isinstance(value, AgentExecutorResponse):
+            return [c for m in value.agent_response.messages for c in m.contents]
+        if isinstance(value, Message):
+            return list(value.contents)
+        if isinstance(value, Content):
+            return [value]
+        return []
+
+    return any(c.type == "function_approval_request" for c in _contents_of(data))
+
+
+async def test_agent_executor_does_not_double_emit_approval_non_streaming() -> None:
+    """Non-streaming: approval payload must only appear in the ``request_info`` event.
+
+    Regression test for the bug where ``AgentExecutor._run_agent`` first
+    ``yield_output``-ed the response (carrying the approval Content) and then
+    additionally emitted a ``request_info`` event for the same payload.
+    """
+    agent = _ApprovalEmittingAgent(id="approve_agent", name="ApproveAgent", approval_request_id="apr_ns_1")
+    executor = AgentExecutor(agent, id="approve_exec")
+    workflow = WorkflowBuilder(start_executor=executor).build()
+
+    request_info_events: list[WorkflowEvent[Any]] = []
+    output_events: list[WorkflowEvent[Any]] = []
+
+    for event in await workflow.run("please delete it"):
+        if event.type == "request_info":
+            request_info_events.append(event)
+        elif event.type == "output":
+            output_events.append(event)
+
+    assert len(request_info_events) == 1
+    assert _has_approval_payload(request_info_events[0])
+    # The approval payload must not also be surfaced as a workflow output.
+    assert not any(_has_approval_payload(e) for e in output_events)
+    assert agent.run_count == 1
+
+
+async def test_agent_executor_does_not_double_emit_approval_streaming() -> None:
+    """Streaming: per-update approval payload must not be ``yield_output``-ed."""
+    agent = _ApprovalEmittingAgent(id="approve_agent_s", name="ApproveAgentS", approval_request_id="apr_st_1")
+    executor = AgentExecutor(agent, id="approve_exec_s")
+    workflow = WorkflowBuilder(start_executor=executor).build()
+
+    request_info_events: list[WorkflowEvent[Any]] = []
+    output_events: list[WorkflowEvent[Any]] = []
+
+    async for event in workflow.run("please delete it", stream=True):
+        if event.type == "request_info":
+            request_info_events.append(event)
+        elif event.type == "output":
+            output_events.append(event)
+
+    assert len(request_info_events) == 1
+    assert _has_approval_payload(request_info_events[0])
+    assert not any(_has_approval_payload(e) for e in output_events)
+    assert agent.run_count == 1
+
+
+async def test_agent_executor_request_info_uses_user_input_request_id() -> None:
+    """``ctx.request_info`` must register the request under the agent's approval id.
+
+    This makes the workflow's pending-request id round-trip with the
+    ``function_approval_response.id`` the caller echoes back, so
+    ``Workflow._send_responses_internal`` can look it up directly.
+    """
+    agent = _ApprovalEmittingAgent(id="approve_agent_id", name="ApproveAgentId", approval_request_id="apr_match")
+    executor = AgentExecutor(agent, id="approve_exec_id")
+    workflow = WorkflowBuilder(start_executor=executor).build()
+
+    request_info_events: list[WorkflowEvent[Any]] = []
+    async for event in workflow.run("please delete it", stream=True):
+        if event.type == "request_info":
+            request_info_events.append(event)
+
+    assert len(request_info_events) == 1
+    assert request_info_events[0].request_id == "apr_match"
+
+
+# endregion Tool approval emission

python/packages/core/tests/workflow/test_workflow_agent.py

@@ -1,6 +1,5 @@
 # Copyright (c) Microsoft. All rights reserved.
 
-import json
 import uuid
 from collections.abc import Awaitable, Sequence
 from dataclasses import dataclass
@@ -30,6 +29,20 @@ from agent_framework import (
     handler,
     response_handler,
 )
+from agent_framework._workflows._typing_utils import deserialize_type
+
+
+@dataclass
+class HandoffRequest:
+    """Module-level dataclass used by request_info tests.
+
+    Defined at module scope (not nested inside a test method) so
+    ``serialize_type``/``deserialize_type`` can round-trip the request_type via
+    the importable qualified name ``tests.workflow.test_workflow_agent.HandoffRequest``.
+    """
+
+    target_agent: str
+    reason: str
 
 
 class SimpleExecutor(Executor):
@@ -240,52 +253,45 @@ class TestWorkflowAgent:
         # Should have received an approval request for the request info
         assert len(updates) > 0
 
-        approval_update: AgentResponseUpdate | None = None
+        request_update: AgentResponseUpdate | None = None
         for update in updates:
-            if any(content.type == "function_approval_request" for content in update.contents):
-                approval_update = update
+            if any(content.type == "function_call" for content in update.contents):
+                request_update = update
                 break
 
-        assert approval_update is not None, "Should have received a request_info approval request"
+        assert request_update is not None, "Should have received a request_info wrapped in a function_call content"
 
-        function_call = next(content for content in approval_update.contents if content.type == "function_call")
-        approval_request = next(
-            content for content in approval_update.contents if content.type == "function_approval_request"
-        )
+        request_function_call = next(content for content in request_update.contents if content.type == "function_call")
+        assert request_function_call.call_id is not None
 
         # Verify the function call has expected structure
-        assert function_call.call_id is not None
-        assert function_call.name == "request_info"
-        assert isinstance(function_call.arguments, dict)
-        assert function_call.arguments.get("request_id") == approval_request.id
+        assert request_function_call.name == WorkflowAgent.REQUEST_INFO_FUNCTION_NAME
+        assert isinstance(request_function_call.arguments, dict)
+        assert request_function_call.arguments.get("request_id") is not None
+        assert request_function_call.arguments.get("request_event") is not None
+        request_event = request_function_call.arguments["request_event"]
+        assert request_event.get("type") == "request_info"
+        assert deserialize_type(request_event.get("response_type")) is str
 
-        # Approval request should reference the same function call
-        assert approval_request.id is not None
-        assert approval_request.function_call is not None
-        assert approval_request.function_call.call_id == function_call.call_id
-        assert approval_request.function_call.name == function_call.name
+        deserialized_args = WorkflowAgent.RequestInfoFunctionArgs.from_dict(request_function_call.arguments)
+        assert deserialized_args.request_id == request_function_call.call_id
+        assert isinstance(deserialized_args.request_event, WorkflowEvent)
+        assert deserialized_args.request_event.type == "request_info"
+        assert deserialized_args.request_event.data == "Mock request data"
+        assert deserialized_args.request_event.response_type is str
 
         # Verify the request is tracked in pending_requests
-        assert len(agent.pending_requests) == 1
-        assert function_call.call_id in agent.pending_requests
+        pending_requests = await workflow._runner_context.get_pending_request_info_events()
+        assert len(pending_requests) == 1
+        assert request_function_call.call_id in pending_requests
 
-        # Now provide an approval response with updated arguments to test continuation
-        response_args = WorkflowAgent.RequestInfoFunctionArgs(
-            request_id=approval_request.id,
-            data="User provided answer",
-        ).to_dict()
-
-        approval_response = Content.from_function_approval_response(
-            approved=True,
-            id=approval_request.id,
-            function_call=Content.from_function_call(
-                call_id=function_call.call_id,
-                name=function_call.name,
-                arguments=response_args,
-            ),
+        # Now provide a function result response with updated arguments to test continuation
+        function_result = Content.from_function_result(
+            call_id=request_function_call.call_id,
+            result="Mock response to request info",
         )
 
-        response_message = Message(role="user", contents=[approval_response])
+        response_message = Message(role="user", contents=[function_result])
 
         # Continue the workflow with the response
         continuation_result = await agent.run(response_message)
@@ -294,16 +300,11 @@ class TestWorkflowAgent:
         assert isinstance(continuation_result, AgentResponse)
 
         # Verify cleanup - pending requests should be cleared after function response handling
-        assert len(agent.pending_requests) == 0
+        pending_requests = await workflow._runner_context.get_pending_request_info_events()
+        assert len(pending_requests) == 0
 
     def test_request_info_dataclass_arguments_are_serialized_when_content_is_created(self) -> None:
         """Test WorkflowAgent prepares request_info arguments before observability captures messages."""
-
-        @dataclass
-        class HandoffRequest:
-            target_agent: str
-            reason: str
-
         executor = SimpleExecutor(id="executor1", response_text="Response")
         workflow = WorkflowBuilder(start_executor=executor).build()
         agent = WorkflowAgent(workflow=workflow, name="Request Test Agent")
@@ -314,14 +315,367 @@ class TestWorkflowAgent:
             response_type=str,
         )
 
-        function_call, approval_request = agent._process_request_info_event(event)  # pyright: ignore[reportPrivateUsage]
+        request_function_call = agent._process_request_info_event(event)  # pyright: ignore[reportPrivateUsage]
 
-        assert function_call.arguments == {
-            "request_id": "request_123",
-            "data": {"target_agent": "helper", "reason": "overflow"},
-        }
-        assert approval_request.function_call is function_call
-        assert json.loads(json.dumps(function_call.arguments)) == function_call.arguments
+        assert request_function_call.call_id == "request_123"
+        assert isinstance(request_function_call.arguments, dict)
+        assert request_function_call.arguments.get("request_event") is not None
+        request_event = request_function_call.arguments["request_event"]
+        assert request_event.get("type") == "request_info"
+        assert request_event.get("request_id") == "request_123"
+        assert request_event.get("source_executor_id") == "executor1"
+        assert deserialize_type(request_event.get("response_type")) is str
+        assert request_event.get("data") == HandoffRequest(target_agent="helper", reason="overflow")
+
+        deserialized_args = WorkflowAgent.RequestInfoFunctionArgs.from_dict(request_function_call.arguments)
+        assert deserialized_args.request_id == "request_123"
+        assert isinstance(deserialized_args.request_event, WorkflowEvent)
+        assert deserialized_args.request_event.type == "request_info"
+        assert deserialized_args.request_event.data == HandoffRequest(target_agent="helper", reason="overflow")
+        assert deserialized_args.request_event.response_type is str
+
+    def test_process_request_info_event_passes_through_function_approval_request(self) -> None:
+        """If the event data is already a function approval request, it is forwarded unchanged.
+
+        Tool-approval requests emitted by an inner agent surface as ``Content``
+        objects with ``user_input_request=True``. ``WorkflowAgent`` must not
+        re-wrap these inside a synthesized ``request_info`` function call;
+        instead it should return the original content as-is so callers can
+        respond with a matching ``function_approval_response``.
+        """
+        executor = SimpleExecutor(id="executor1", response_text="Response")
+        workflow = WorkflowBuilder(start_executor=executor).build()
+        agent = WorkflowAgent(workflow=workflow, name="Approval Passthrough Agent")
+
+        approval_id = "approval-passthrough-1"
+        inner_function_call = Content.from_function_call(
+            call_id="tool-call-1",
+            name="delete_file",
+            arguments={"path": "/tmp/x"},
+        )
+        approval_request = Content.from_function_approval_request(
+            id=approval_id,
+            function_call=inner_function_call,
+        )
+        event = WorkflowEvent.request_info(
+            request_id=approval_id,
+            source_executor_id="executor1",
+            request_data=approval_request,
+            response_type=Content,
+        )
+
+        result = agent._process_request_info_event(event)  # pyright: ignore[reportPrivateUsage]
+
+        # The original FunctionApprovalRequestContent is returned as-is — same
+        # instance, with the original tool name preserved (NOT replaced by the
+        # synthesized REQUEST_INFO_FUNCTION_NAME).
+        assert result is approval_request
+        assert result.type == "function_approval_request"
+        assert result.id == approval_id
+        assert result.user_input_request is True
+        assert result.function_call is inner_function_call  # type: ignore[attr-defined]
+        assert result.function_call.name == "delete_file"  # type: ignore[attr-defined]
+        assert result.function_call.name != WorkflowAgent.REQUEST_INFO_FUNCTION_NAME  # type: ignore[attr-defined]
+
+    def test_extract_function_responses_passes_through_approval_response_approved(self) -> None:
+        """A function_approval_response with approved=True is keyed by content.id and forwarded as-is.
+
+        After the refactor, ``WorkflowAgent`` no longer unwraps a synthesized
+        ``request_info`` function call from approval responses — the response
+        content is routed straight back to the workflow under its own ``id``,
+        which matches the pending request id surfaced by
+        ``_process_request_info_event``.
+        """
+        executor = SimpleExecutor(id="executor1", response_text="Response")
+        workflow = WorkflowBuilder(start_executor=executor).build()
+        agent = WorkflowAgent(workflow=workflow, name="Approval Response Agent")
+
+        approval_id = "approval-response-approved-1"
+        inner_function_call = Content.from_function_call(
+            call_id="tool-call-1",
+            name="delete_file",
+            arguments={"path": "/tmp/x"},
+        )
+        approval_request = Content.from_function_approval_request(
+            id=approval_id,
+            function_call=inner_function_call,
+        )
+        approval_response = approval_request.to_function_approval_response(approved=True)  # type: ignore[attr-defined]
+        message = Message(role="user", contents=[approval_response])
+
+        responses = agent._extract_function_responses([message])  # pyright: ignore[reportPrivateUsage]
+
+        assert set(responses.keys()) == {approval_id}
+        assert responses[approval_id] is approval_response
+        assert responses[approval_id].approved is True  # type: ignore[attr-defined]
+
+    def test_extract_function_responses_passes_through_approval_response_denied(self) -> None:
+        """A function_approval_response with approved=False is forwarded the same way as an approval.
+
+        Only the ``approved`` flag changes — routing back to the workflow is
+        identical for accept and reject paths.
+        """
+        executor = SimpleExecutor(id="executor1", response_text="Response")
+        workflow = WorkflowBuilder(start_executor=executor).build()
+        agent = WorkflowAgent(workflow=workflow, name="Approval Response Agent")
+
+        approval_id = "approval-response-denied-1"
+        inner_function_call = Content.from_function_call(
+            call_id="tool-call-2",
+            name="send_email",
+            arguments={"to": "alice@example.com"},
+        )
+        approval_request = Content.from_function_approval_request(
+            id=approval_id,
+            function_call=inner_function_call,
+        )
+        approval_response = approval_request.to_function_approval_response(approved=False)  # type: ignore[attr-defined]
+        message = Message(role="user", contents=[approval_response])
+
+        responses = agent._extract_function_responses([message])  # pyright: ignore[reportPrivateUsage]
+
+        assert set(responses.keys()) == {approval_id}
+        assert responses[approval_id] is approval_response
+        assert responses[approval_id].approved is False  # type: ignore[attr-defined]
+
+    async def test_function_approval_request_flows_end_to_end_approved(self) -> None:
+        """End-to-end: an executor emits a function_approval_request, the agent
+        forwards it unchanged, and an ``approved=True`` response resumes the workflow.
+
+        This exercises the full pass-through path:
+        ``ctx.request_info(approval_content, ...)`` -> ``WorkflowAgent`` surfaces
+        the original ``FunctionApprovalRequestContent`` -> caller responds with a
+        ``FunctionApprovalResponseContent`` -> ``WorkflowAgent`` routes it back
+        to the workflow which delivers it to the executor's ``@response_handler``.
+        """
+        approval_id = "e2e-approval-1"
+        inner_function_call = Content.from_function_call(
+            call_id="tool-call-e2e-1",
+            name="delete_file",
+            arguments={"path": "/tmp/x"},
+        )
+        approval_request = Content.from_function_approval_request(
+            id=approval_id,
+            function_call=inner_function_call,
+        )
+
+        class ApprovalRequestingExecutor(Executor):
+            @handler
+            async def handle_message(self, _: list[Message], ctx: WorkflowContext) -> None:
+                await ctx.request_info(approval_request, Content, request_id=approval_id)
+
+            @response_handler
+            async def handle_response(
+                self,
+                original_request: Content,
+                response: Content,
+                ctx: WorkflowContext[Never, AgentResponse],
+            ) -> None:
+                assert response.type == "function_approval_response"
+                assert response.id == approval_id  # type: ignore[attr-defined]
+                approved = bool(response.approved)  # type: ignore[attr-defined]
+                tool_name = original_request.function_call.name  # type: ignore[attr-defined]
+                await ctx.yield_output(
+                    AgentResponse(
+                        messages=[
+                            Message(
+                                role="assistant",
+                                contents=[Content.from_text(text=f"{tool_name} approved={approved}")],
+                            )
+                        ]
+                    )
+                )
+
+        executor = ApprovalRequestingExecutor(id="approval_requester")
+        workflow = WorkflowBuilder(start_executor=executor).build()
+        agent = WorkflowAgent(workflow=workflow, name="E2E Approval Agent")
+
+        # First run: workflow pauses with the approval request.
+        first = await agent.run("please delete it")
+        assert isinstance(first, AgentResponse)
+
+        forwarded = next(
+            (
+                c
+                for m in first.messages
+                for c in m.contents
+                if c.type == "function_approval_request" and c.id == approval_id
+            ),
+            None,
+        )
+        assert forwarded is approval_request, "Approval request must surface unchanged"
+
+        pending = await workflow._runner_context.get_pending_request_info_events()
+        assert approval_id in pending
+
+        # Respond with approved=True.
+        approval_response = approval_request.to_function_approval_response(approved=True)  # type: ignore[attr-defined]
+        final = await agent.run(Message(role="user", contents=[approval_response]))
+
+        assert isinstance(final, AgentResponse)
+        final_text = " ".join(m.text or "" for m in final.messages)
+        assert "delete_file approved=True" in final_text
+
+        pending = await workflow._runner_context.get_pending_request_info_events()
+        assert approval_id not in pending
+
+    async def test_function_approval_request_flows_end_to_end_denied(self) -> None:
+        """End-to-end denied path: ``approved=False`` is delivered to the executor's
+        response handler so the workflow can branch on the rejection."""
+        approval_id = "e2e-approval-deny-1"
+        inner_function_call = Content.from_function_call(
+            call_id="tool-call-e2e-deny-1",
+            name="send_email",
+            arguments={"to": "alice@example.com"},
+        )
+        approval_request = Content.from_function_approval_request(
+            id=approval_id,
+            function_call=inner_function_call,
+        )
+
+        class ApprovalRequestingExecutor(Executor):
+            @handler
+            async def handle_message(self, _: list[Message], ctx: WorkflowContext) -> None:
+                await ctx.request_info(approval_request, Content, request_id=approval_id)
+
+            @response_handler
+            async def handle_response(
+                self,
+                original_request: Content,
+                response: Content,
+                ctx: WorkflowContext[Never, AgentResponse],
+            ) -> None:
+                assert response.type == "function_approval_response"
+                assert response.id == approval_id  # type: ignore[attr-defined]
+                approved = bool(response.approved)  # type: ignore[attr-defined]
+                tool_name = original_request.function_call.name  # type: ignore[attr-defined]
+                await ctx.yield_output(
+                    AgentResponse(
+                        messages=[
+                            Message(
+                                role="assistant",
+                                contents=[Content.from_text(text=f"{tool_name} approved={approved}")],
+                            )
+                        ]
+                    )
+                )
+
+        executor = ApprovalRequestingExecutor(id="approval_requester_deny")
+        workflow = WorkflowBuilder(start_executor=executor).build()
+        agent = WorkflowAgent(workflow=workflow, name="E2E Approval Deny Agent")
+
+        first = await agent.run("please send")
+        assert isinstance(first, AgentResponse)
+        forwarded = next(
+            (
+                c
+                for m in first.messages
+                for c in m.contents
+                if c.type == "function_approval_request" and c.id == approval_id
+            ),
+            None,
+        )
+        assert forwarded is approval_request
+
+        # Respond with approved=False.
+        approval_response = approval_request.to_function_approval_response(approved=False)  # type: ignore[attr-defined]
+        final = await agent.run(Message(role="user", contents=[approval_response]))
+
+        assert isinstance(final, AgentResponse)
+        final_text = " ".join(m.text or "" for m in final.messages)
+        assert "send_email approved=False" in final_text
+
+        pending = await workflow._runner_context.get_pending_request_info_events()
+        assert approval_id not in pending
+
+    async def test_request_info_non_approval_flows_end_to_end(self) -> None:
+        """End-to-end: when request data is not a function approval content, the
+        agent surfaces a synthesized ``function_call`` (name=REQUEST_INFO_FUNCTION_NAME)
+        and routes a matching ``function_result`` back to the executor.
+        """
+        captured: dict[str, Any] = {}
+
+        class HandoffRequestingExecutor(Executor):
+            @handler
+            async def handle_message(self, _: list[Message], ctx: WorkflowContext) -> None:
+                await ctx.request_info(
+                    HandoffRequest(target_agent="helper", reason="overflow"),
+                    str,
+                )
+
+            @response_handler
+            async def handle_response(
+                self,
+                original_request: HandoffRequest,
+                response: str,
+                ctx: WorkflowContext[Never, AgentResponse],
+            ) -> None:
+                captured["original"] = original_request
+                captured["response"] = response
+                await ctx.yield_output(
+                    AgentResponse(
+                        messages=[
+                            Message(
+                                role="assistant",
+                                contents=[
+                                    Content.from_text(text=f"handoff to {original_request.target_agent}: {response}")
+                                ],
+                            )
+                        ]
+                    )
+                )
+
+        executor = HandoffRequestingExecutor(id="handoff_requester")
+        workflow = WorkflowBuilder(start_executor=executor).build()
+        agent = WorkflowAgent(workflow=workflow, name="E2E Handoff Agent")
+
+        # First run: workflow pauses with a synthesized request_info function_call.
+        first = await agent.run("start handoff")
+        assert isinstance(first, AgentResponse)
+
+        function_call = next(
+            (
+                c
+                for m in first.messages
+                for c in m.contents
+                if c.type == "function_call" and c.name == WorkflowAgent.REQUEST_INFO_FUNCTION_NAME
+            ),
+            None,
+        )
+        assert function_call is not None, "Expected a synthesized request_info function_call"
+        assert function_call.call_id is not None
+        assert isinstance(function_call.arguments, dict)
+        request_id = function_call.arguments["request_id"]
+        assert function_call.call_id == request_id
+        request_payload = function_call.arguments["request_event"]
+        assert request_payload.get("type") == "request_info"
+        assert request_payload.get("data") == HandoffRequest(target_agent="helper", reason="overflow")
+
+        deserialized_args = WorkflowAgent.RequestInfoFunctionArgs.from_dict(function_call.arguments)
+        assert deserialized_args.request_id == request_id
+        assert isinstance(deserialized_args.request_event, WorkflowEvent)
+        assert deserialized_args.request_event.type == "request_info"
+        assert deserialized_args.request_event.data == HandoffRequest(target_agent="helper", reason="overflow")
+        assert deserialized_args.request_event.response_type is str
+
+        pending = await workflow._runner_context.get_pending_request_info_events()
+        assert request_id in pending
+
+        # Respond with a function_result keyed by the call_id.
+        function_result = Content.from_function_result(call_id=request_id, result="ok-do-it")
+        final = await agent.run(Message(role="user", contents=[function_result]))
+
+        assert isinstance(final, AgentResponse)
+        final_text = " ".join(m.text or "" for m in final.messages)
+        assert "handoff to helper: ok-do-it" in final_text
+
+        # The executor's response handler received the original request and the response.
+        assert isinstance(captured.get("original"), HandoffRequest)
+        assert captured["original"].target_agent == "helper"
+        assert captured["response"] == "ok-do-it"
+
+        pending = await workflow._runner_context.get_pending_request_info_events()
+        assert request_id not in pending
 
     def test_workflow_as_agent_method(self) -> None:
         """Test that Workflow.as_agent() creates a properly configured WorkflowAgent."""
@@ -1592,3 +1946,406 @@ class TestWorkflowAgentMergeUpdates:
 
         # Order: text (user), text (assistant), function_result (orphan at end)
         assert content_types == ["text", "text", "function_result"]
+
+
+class _ToolApprovalMockAgent(SupportsAgentRun):
+    """Mock agent whose first run returns a FunctionApprovalRequestContent.
+
+    Subsequent runs (after receiving an approval response in the input messages)
+    return a final assistant text response that echoes the approved arguments.
+
+    This mirrors a real agent whose tool invocation requires user approval.
+    """
+
+    def __init__(
+        self,
+        name: str,
+        *,
+        tool_name: str = "delete_file",
+        tool_arguments: dict[str, Any] | None = None,
+        approval_request_ids: Sequence[str] | None = None,
+    ) -> None:
+        self.id = str(uuid.uuid4())
+        self.name = name
+        self.description: str | None = None
+        self._tool_name = tool_name
+        self._tool_arguments = tool_arguments or {"path": "/tmp/example"}
+        # Pre-allocated request ids so the test can verify what the WorkflowAgent forwards.
+        self._approval_request_ids: list[str] = list(approval_request_ids) if approval_request_ids else []
+        self.run_count = 0
+        # Inputs received on the most recent (continuation) run, for assertions.
+        self.last_run_messages: list[Message] = []
+
+    def create_session(self, **kwargs: Any) -> AgentSession:
+        return AgentSession()
+
+    def get_session(self, *, service_session_id: str, **kwargs: Any) -> AgentSession:
+        return AgentSession()
+
+    def _next_request_id(self) -> str:
+        if self._approval_request_ids:
+            return self._approval_request_ids.pop(0)
+        return str(uuid.uuid4())
+
+    def _build_approval_request(self) -> Content:
+        request_id = self._next_request_id()
+        function_call = Content.from_function_call(
+            call_id=request_id,
+            name=self._tool_name,
+            arguments=self._tool_arguments,
+        )
+        return Content.from_function_approval_request(id=request_id, function_call=function_call)
+
+    @overload
+    def run(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None = ...,
+        *,
+        stream: Literal[False] = ...,
+        session: AgentSession | None = ...,
+        **kwargs: Any,
+    ) -> Awaitable[AgentResponse[Any]]: ...
+
+    @overload
+    def run(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None = ...,
+        *,
+        stream: Literal[True],
+        session: AgentSession | None = ...,
+        **kwargs: Any,
+    ) -> ResponseStream[AgentResponseUpdate, AgentResponse[Any]]: ...
+
+    def run(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None = None,
+        *,
+        stream: bool = False,
+        session: AgentSession | None = None,
+        **kwargs: Any,
+    ) -> Awaitable[AgentResponse] | ResponseStream[AgentResponseUpdate, AgentResponse]:
+        if stream:
+            return self._run_stream(messages=messages, session=session, **kwargs)
+        return self._run(messages=messages, session=session, **kwargs)
+
+    def _normalize(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None,
+    ) -> list[Message]:
+        if messages is None:
+            return []
+        if isinstance(messages, str):
+            return [Message(role="user", contents=[Content.from_text(text=messages)])]
+        if isinstance(messages, Message):
+            return [messages]
+        if isinstance(messages, Content):
+            return [Message(role="user", contents=[messages])]
+        result: list[Message] = []
+        for item in messages:
+            if isinstance(item, Message):
+                result.append(item)
+            elif isinstance(item, Content):
+                result.append(Message(role="user", contents=[item]))
+            else:
+                result.append(Message(role="user", contents=[Content.from_text(text=item)]))
+        return result
+
+    def _approval_responses_in(self, messages: list[Message]) -> list[Content]:
+        approvals: list[Content] = []
+        for msg in messages:
+            for content in msg.contents:
+                if content.type == "function_approval_response":
+                    approvals.append(content)
+        return approvals
+
+    async def _run(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None = None,
+        *,
+        session: AgentSession | None = None,
+        **kwargs: Any,
+    ) -> AgentResponse:
+        normalized = self._normalize(messages)
+        self.last_run_messages = normalized
+        self.run_count += 1
+
+        approvals = self._approval_responses_in(normalized)
+        if approvals:
+            # Continuation: reflect approved arguments in the final response text.
+            approved_text = "; ".join(
+                f"approved={a.approved} id={a.id}"  # type: ignore[attr-defined]
+                for a in approvals
+            )
+            return AgentResponse(messages=[Message("assistant", [Content.from_text(text=f"done ({approved_text})")])])
+
+        # First run: ask for tool approval.
+        approval = self._build_approval_request()
+        return AgentResponse(messages=[Message("assistant", [approval])])
+
+    def _run_stream(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None = None,
+        *,
+        session: AgentSession | None = None,
+        **kwargs: Any,
+    ) -> ResponseStream[AgentResponseUpdate, AgentResponse]:
+        normalized = self._normalize(messages)
+        self.last_run_messages = normalized
+        self.run_count += 1
+        approvals = self._approval_responses_in(normalized)
+
+        async def _iter():
+            if approvals:
+                approved_text = "; ".join(
+                    f"approved={a.approved} id={a.id}"  # type: ignore[attr-defined]
+                    for a in approvals
+                )
+                yield AgentResponseUpdate(
+                    contents=[Content.from_text(text=f"done ({approved_text})")],
+                    role="assistant",
+                    author_name=self.name,
+                )
+                return
+            approval = self._build_approval_request()
+            yield AgentResponseUpdate(
+                contents=[approval],
+                role="assistant",
+                author_name=self.name,
+            )
+
+        return ResponseStream(_iter(), finalizer=AgentResponse.from_updates)
+
+
+class TestWorkflowAgentToolApproval:
+    """Tests for tool-approval requests bubbling through WorkflowAgent.
+
+    Covers the case where a workflow contains an AgentExecutor whose underlying
+    agent emits a FunctionApprovalRequestContent (tool needing user approval).
+    The WorkflowAgent must:
+      * forward the original FunctionApprovalRequestContent unchanged (no
+        wrapping inside a synthesized 'request_info' function call), and
+      * route a subsequent FunctionApprovalResponseContent back to the
+        AgentExecutor so the agent can resume.
+    """
+
+    def _find_approval_request(
+        self,
+        contents: Sequence[Content],
+        tool_name: str,
+    ) -> Content | None:
+        for content in contents:
+            if (
+                content.type == "function_approval_request"
+                and getattr(content.function_call, "name", None) == tool_name  # type: ignore[attr-defined]
+            ):
+                return content
+        return None
+
+    async def test_tool_approval_request_forwarded_unchanged(self) -> None:
+        """The agent's FunctionApprovalRequestContent surfaces verbatim (not re-wrapped)."""
+        approval_id = "approval-abc-123"
+        mock_agent = _ToolApprovalMockAgent(
+            name="approval-agent",
+            tool_name="delete_file",
+            tool_arguments={"path": "/tmp/secret.txt"},
+            approval_request_ids=[approval_id],
+        )
+
+        @executor
+        async def start(messages: list[Message], ctx: WorkflowContext[AgentExecutorRequest]) -> None:
+            await ctx.send_message(AgentExecutorRequest(messages=messages, should_respond=True))
+
+        workflow = WorkflowBuilder(start_executor=start).add_edge(start, mock_agent).build()
+        agent = WorkflowAgent(workflow=workflow, name="Approval Test Agent")
+
+        result = await agent.run("please delete the file")
+
+        assert isinstance(result, AgentResponse)
+
+        # Locate the approval request emitted by the WorkflowAgent.
+        all_contents: list[Content] = [c for m in result.messages for c in m.contents]
+        approval = self._find_approval_request(all_contents, tool_name="delete_file")
+        assert approval is not None, "WorkflowAgent did not forward the tool approval request"
+
+        # The id and inner function_call must match what the underlying agent produced
+        # — i.e. the WorkflowAgent must NOT have re-wrapped it inside a synthesized
+        # 'request_info' approval request.
+        assert approval.id == approval_id
+        function_call = approval.function_call  # type: ignore[attr-defined]
+        assert function_call is not None
+        assert function_call.name == "delete_file"
+        assert function_call.name != WorkflowAgent.REQUEST_INFO_FUNCTION_NAME
+        assert function_call.arguments == {"path": "/tmp/secret.txt"}
+
+        # The agent must be paused awaiting the approval response.
+        pending = await workflow._runner_context.get_pending_request_info_events()
+        assert approval_id in pending
+
+    async def test_tool_approval_request_forwarded_unchanged_streaming(self) -> None:
+        """Streaming variant: the approval request is forwarded as-is in updates."""
+        approval_id = "approval-stream-1"
+        mock_agent = _ToolApprovalMockAgent(
+            name="approval-agent-stream",
+            tool_name="send_email",
+            tool_arguments={"to": "alice@example.com"},
+            approval_request_ids=[approval_id],
+        )
+
+        @executor
+        async def start(messages: list[Message], ctx: WorkflowContext[AgentExecutorRequest]) -> None:
+            await ctx.send_message(AgentExecutorRequest(messages=messages, should_respond=True))
+
+        workflow = WorkflowBuilder(start_executor=start).add_edge(start, mock_agent).build()
+        agent = WorkflowAgent(workflow=workflow, name="Approval Stream Agent")
+
+        updates: list[AgentResponseUpdate] = []
+        async for update in agent.run("hi", stream=True):
+            updates.append(update)
+
+        approval_updates = [u for u in updates if any(c.type == "function_approval_request" for c in u.contents)]
+        assert approval_updates, "Streaming did not surface a tool approval request"
+
+        approval = self._find_approval_request(approval_updates[-1].contents, tool_name="send_email")
+        assert approval is not None
+        assert approval.id == approval_id
+        function_call = approval.function_call  # type: ignore[attr-defined]
+        assert function_call is not None
+        assert function_call.name == "send_email"
+        assert function_call.name != WorkflowAgent.REQUEST_INFO_FUNCTION_NAME
+        assert function_call.arguments == {"to": "alice@example.com"}
+
+    async def test_tool_approval_response_resumes_agent(self) -> None:
+        """Sending the approval response back resumes the agent and clears pending requests."""
+        approval_id = "approval-resume-1"
+        mock_agent = _ToolApprovalMockAgent(
+            name="approval-resume-agent",
+            tool_name="delete_file",
+            tool_arguments={"path": "/tmp/x"},
+            approval_request_ids=[approval_id],
+        )
+
+        @executor
+        async def start(messages: list[Message], ctx: WorkflowContext[AgentExecutorRequest]) -> None:
+            await ctx.send_message(AgentExecutorRequest(messages=messages, should_respond=True))
+
+        workflow = WorkflowBuilder(start_executor=start).add_edge(start, mock_agent).build()
+        agent = WorkflowAgent(workflow=workflow, name="Approval Resume Agent")
+
+        first_result = await agent.run("delete it")
+        approval = self._find_approval_request(
+            [c for m in first_result.messages for c in m.contents],
+            tool_name="delete_file",
+        )
+        assert approval is not None
+        assert mock_agent.run_count == 1
+
+        # Build the approval response. NOTE: the inner function_call's name is the
+        # original tool name ('delete_file'), NOT 'request_info'. This exercises the
+        # branch in WorkflowAgent._extract_function_responses that routes raw
+        # tool-approval responses straight through using content.id.
+        approval_response = approval.to_function_approval_response(approved=True)  # type: ignore[attr-defined]
+        response_message = Message(role="user", contents=[approval_response])
+
+        final_result = await agent.run(response_message)
+        assert isinstance(final_result, AgentResponse)
+
+        # The mock agent should have been invoked a second time and seen the
+        # approval response in its inputs.
+        assert mock_agent.run_count == 2
+        approvals_seen = [
+            c for m in mock_agent.last_run_messages for c in m.contents if c.type == "function_approval_response"
+        ]
+        assert len(approvals_seen) == 1
+        assert approvals_seen[0].id == approval_id  # type: ignore[attr-defined]
+        assert approvals_seen[0].approved is True  # type: ignore[attr-defined]
+
+        # The pending approval should now be cleared.
+        pending = await workflow._runner_context.get_pending_request_info_events()
+        assert approval_id not in pending
+
+        # The final assistant message reflects the resumption.
+        final_text = " ".join(m.text or "" for m in final_result.messages)
+        assert "done" in final_text
+        assert approval_id in final_text
+
+    async def test_tool_approval_response_rejected_resumes_agent(self) -> None:
+        """Rejection path: ``approved=False`` is forwarded to the inner agent and clears the pending request.
+
+        The WorkflowAgent must route a rejection response back to the paused
+        ``AgentExecutor`` exactly the same way as an approval — only the
+        ``approved`` flag differs. The inner agent decides what to do with it.
+        """
+        approval_id = "approval-reject-1"
+        mock_agent = _ToolApprovalMockAgent(
+            name="approval-reject-agent",
+            tool_name="delete_file",
+            tool_arguments={"path": "/tmp/x"},
+            approval_request_ids=[approval_id],
+        )
+
+        @executor
+        async def start(messages: list[Message], ctx: WorkflowContext[AgentExecutorRequest]) -> None:
+            await ctx.send_message(AgentExecutorRequest(messages=messages, should_respond=True))
+
+        workflow = WorkflowBuilder(start_executor=start).add_edge(start, mock_agent).build()
+        agent = WorkflowAgent(workflow=workflow, name="Approval Reject Agent")
+
+        first_result = await agent.run("delete it")
+        approval = self._find_approval_request(
+            [c for m in first_result.messages for c in m.contents],
+            tool_name="delete_file",
+        )
+        assert approval is not None
+        assert mock_agent.run_count == 1
+
+        # Reject the tool invocation.
+        approval_response = approval.to_function_approval_response(approved=False)  # type: ignore[attr-defined]
+        response_message = Message(role="user", contents=[approval_response])
+
+        final_result = await agent.run(response_message)
+        assert isinstance(final_result, AgentResponse)
+
+        # The inner agent must have been resumed and seen ``approved=False``.
+        assert mock_agent.run_count == 2
+        approvals_seen = [
+            c for m in mock_agent.last_run_messages for c in m.contents if c.type == "function_approval_response"
+        ]
+        assert len(approvals_seen) == 1
+        assert approvals_seen[0].id == approval_id  # type: ignore[attr-defined]
+        assert approvals_seen[0].approved is False  # type: ignore[attr-defined]
+
+        # Pending approval cleared regardless of approve/reject.
+        pending = await workflow._runner_context.get_pending_request_info_events()
+        assert approval_id not in pending
+
+        # The final assistant message reflects the rejection.
+        final_text = " ".join(m.text or "" for m in final_result.messages)
+        assert "approved=False" in final_text
+        assert approval_id in final_text
+
+    async def test_tool_approval_request_id_matches_pending_request(self) -> None:
+        """The approval request id surfaced by WorkflowAgent matches the workflow's pending request id.
+
+        This guards the AgentExecutor change that forwards
+        request_id=user_input_request.id to ctx.request_info(...), which is what
+        allows the response routed back via WorkflowAgent to resolve the pending
+        request without an id-mismatch error.
+        """
+        approval_id = "approval-id-match-1"
+        mock_agent = _ToolApprovalMockAgent(
+            name="approval-id-match-agent",
+            approval_request_ids=[approval_id],
+        )
+
+        @executor
+        async def start(messages: list[Message], ctx: WorkflowContext[AgentExecutorRequest]) -> None:
+            await ctx.send_message(AgentExecutorRequest(messages=messages, should_respond=True))
+
+        workflow = WorkflowBuilder(start_executor=start).add_edge(start, mock_agent).build()
+        agent = WorkflowAgent(workflow=workflow, name="Approval Id Agent")
+
+        await agent.run("go")
+
+        pending = await workflow._runner_context.get_pending_request_info_events()
+        # The agent's approval id is used as the workflow's pending request id.
+        assert list(pending.keys()) == [approval_id]

python/packages/core/tests/workflow/test_workflow_status.py

@@ -0,0 +1,149 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+"""Tests for the ``Workflow.status`` property."""
+
+from dataclasses import dataclass
+
+import pytest
+
+from agent_framework import (
+    Executor,
+    Workflow,
+    WorkflowBuilder,
+    WorkflowContext,
+    WorkflowEvent,
+    WorkflowRunState,
+    handler,
+    response_handler,
+)
+from agent_framework._workflows._executor import Executor as _Executor
+from agent_framework._workflows._request_info_mixin import RequestInfoMixin
+
+
+class PassThroughExecutor(Executor):
+    """Executor that yields its input as a workflow output and stops."""
+
+    @handler
+    async def passthrough(self, msg: str, ctx: WorkflowContext[str, str]) -> None:
+        await ctx.yield_output(msg)
+
+
+class FailingExecutor(Executor):
+    """Executor that raises at runtime to drive the FAILED status."""
+
+    @handler
+    async def fail(self, msg: int, ctx: WorkflowContext) -> None:  # pragma: no cover - invoked via workflow
+        raise RuntimeError("boom")
+
+
+@dataclass
+class _ApprovalRequest:
+    prompt: str
+    request_id: str = ""
+
+    def __post_init__(self) -> None:
+        if not self.request_id:
+            import uuid
+
+            self.request_id = str(uuid.uuid4())
+
+
+class ApprovalExecutor(_Executor, RequestInfoMixin):
+    """Executor that issues a single request_info call and finalizes on response."""
+
+    def __init__(self, id: str = "approval"):
+        super().__init__(id=id)
+
+    @handler
+    async def start(self, message: str, ctx: WorkflowContext[str, str]) -> None:
+        await ctx.request_info(_ApprovalRequest(prompt=message), bool)
+
+    @response_handler
+    async def on_response(
+        self, original_request: _ApprovalRequest, approved: bool, ctx: WorkflowContext[str, str]
+    ) -> None:
+        await ctx.yield_output(f"approved={approved}")
+
+
+def _build_passthrough_workflow() -> Workflow:
+    executor = PassThroughExecutor(id="p")
+    return WorkflowBuilder(start_executor=executor, output_from=[executor]).build()
+
+
+def _build_failing_workflow() -> Workflow:
+    # FailingExecutor has no workflow_output_types, so we leave designation
+    # implicit; the deprecation warning is filtered at call sites that need it.
+    return WorkflowBuilder(start_executor=FailingExecutor(id="f")).build()
+
+
+def _build_approval_workflow() -> Workflow:
+    executor = ApprovalExecutor(id="approval")
+    return WorkflowBuilder(start_executor=executor, output_from=[executor]).build()
+
+
+async def test_status_default_is_idle_before_first_run():
+    wf = _build_passthrough_workflow()
+    assert wf.status is WorkflowRunState.IDLE
+
+
+async def test_status_is_idle_after_successful_run():
+    wf = _build_passthrough_workflow()
+    await wf.run("hello")
+    assert wf.status is WorkflowRunState.IDLE
+
+
+async def test_status_is_failed_after_failure():
+    wf = _build_failing_workflow()
+    with pytest.raises(RuntimeError, match="boom"):
+        await wf.run(0)
+    assert wf.status is WorkflowRunState.FAILED
+
+
+async def test_status_transitions_during_streaming_run():
+    """Workflow.status mirrors the most recent emitted status event."""
+    wf = _build_passthrough_workflow()
+    observed: list[WorkflowRunState] = []
+
+    async for event in wf.run("hi", stream=True):
+        if isinstance(event, WorkflowEvent) and event.type == "status":
+            # By the time a status event surfaces to the consumer, the property
+            # must already reflect that state (updated in lockstep with emission).
+            assert wf.status == event.state
+            observed.append(event.state)  # type: ignore
+
+    # IN_PROGRESS must precede IDLE; both must appear.
+    assert WorkflowRunState.IN_PROGRESS in observed
+    assert observed[-1] is WorkflowRunState.IDLE
+    assert wf.status is WorkflowRunState.IDLE
+
+
+async def test_status_idle_with_pending_requests_then_resolves_to_idle():
+    wf = _build_approval_workflow()
+
+    request_event: WorkflowEvent | None = None
+    async for event in wf.run("please approve", stream=True):
+        if isinstance(event, WorkflowEvent) and event.type == "request_info":
+            request_event = event
+
+    assert request_event is not None
+    assert wf.status is WorkflowRunState.IDLE_WITH_PENDING_REQUESTS
+
+    async for _ in wf.run(stream=True, responses={request_event.request_id: True}):
+        pass
+
+    assert wf.status is WorkflowRunState.IDLE
+
+
+async def test_status_in_progress_pending_requests_observed_mid_run():
+    """While streaming, status reaches IN_PROGRESS_PENDING_REQUESTS after a request_info event."""
+    wf = _build_approval_workflow()
+    seen_states: list[WorkflowRunState] = []
+
+    async for event in wf.run("please approve", stream=True):
+        if isinstance(event, WorkflowEvent) and event.type == "status":
+            seen_states.append(event.state)  # type: ignore
+
+    assert WorkflowRunState.IN_PROGRESS in seen_states
+    assert WorkflowRunState.IN_PROGRESS_PENDING_REQUESTS in seen_states
+    assert seen_states[-1] is WorkflowRunState.IDLE_WITH_PENDING_REQUESTS
+    assert wf.status is WorkflowRunState.IDLE_WITH_PENDING_REQUESTS

python/packages/foundry/pyproject.toml

@@ -26,7 +26,7 @@ dependencies = [
     "agent-framework-core>=1.8.0,<2",
     "agent-framework-openai>=1.8.0,<2",
     "azure-ai-inference>=1.0.0b9,<1.0.0b10",
-    "azure-ai-projects>=2.1.0,<3.0",
+    "azure-ai-projects>=2.2.0,<3.0",
 ]
 
 [tool.uv]

python/packages/foundry_hosting/agent_framework_foundry_hosting/_responses.py

@@ -567,7 +567,7 @@ class ResponsesHostServer(ResponsesAgentServerHost):
         by the hosting infrastructure or files will be preserved upon deactivation.
         """
         input_items = await context.get_input_items()
-        input_messages = await _items_to_messages(input_items)
+        input_messages = await _items_to_messages(input_items, approval_storage=self._approval_storage)
         is_streaming_request = request.stream is not None and request.stream is True
 
         _, are_options_set = _to_chat_options(request)
@@ -664,7 +664,11 @@ class ResponsesHostServer(ResponsesAgentServerHost):
                 checkpoint_storage=write_storage,
             )
 
-            async for item in _to_outputs_for_messages(response_event_stream, response.messages):
+            async for item in _to_outputs_for_messages(
+                response_event_stream,
+                response.messages,
+                approval_storage=self._approval_storage,
+            ):
                 yield item
 
             await self._delete_not_latest_checkpoints(write_storage, self._agent.workflow.name)
@@ -685,7 +689,9 @@ class ResponsesHostServer(ResponsesAgentServerHost):
                 for event in tracker.handle(content):
                     yield event
                 if tracker.needs_async:
-                    async for item in _to_outputs(response_event_stream, content):
+                    async for item in _to_outputs(
+                        response_event_stream, content, approval_storage=self._approval_storage
+                    ):
                         yield item
                     tracker.needs_async = False
 

python/packages/foundry_hosting/tests/test_responses.py

@@ -11,24 +11,33 @@ the registered _handle_create handler.
 from __future__ import annotations
 
 import json
-from collections.abc import AsyncIterator, Callable
+import uuid
+from collections.abc import AsyncIterator, Awaitable, Callable, Sequence
 from dataclasses import dataclass
+from typing import Literal, overload
 from unittest.mock import AsyncMock, MagicMock, patch
 
 import httpx
 import pytest
 from agent_framework import (
+    AgentExecutorRequest,
     AgentResponse,
     AgentResponseUpdate,
+    AgentSession,
     Content,
     FileCheckpointStorage,
     HistoryProvider,
     Message,
     RawAgent,
     ResponseStream,
+    SupportsAgentRun,
+    WorkflowAgent,
+    WorkflowBuilder,
     WorkflowCheckpoint,
     WorkflowCheckpointException,
+    WorkflowContext,
     WorkflowMessage,
+    executor,
 )
 from azure.ai.agentserver.responses import InMemoryResponseProvider
 from mcp import McpError
@@ -102,7 +111,7 @@ def _make_agent(
     return agent
 
 
-def _make_server(agent: MagicMock, **kwargs: Any) -> ResponsesHostServer:
+def _make_server(agent: Any, **kwargs: Any) -> ResponsesHostServer:
     """Create a ResponsesHostServer with an in-memory store."""
     return ResponsesHostServer(agent, store=InMemoryResponseProvider(), **kwargs)
 
@@ -3469,3 +3478,498 @@ class TestOAuthConsentSurfacing:
 
 
 # endregion
+
+# region Workflow agent hosting (end-to-end)
+
+
+class _ToolApprovalWorkflowAgentMock(SupportsAgentRun):
+    """Inner agent for a hosted ``WorkflowAgent`` whose first run emits a
+    ``FunctionApprovalRequestContent`` and whose follow-up run (after
+    receiving a ``FunctionApprovalResponseContent`` in its inputs) returns a
+    final assistant text response.
+
+    Mirrors a real agent whose tool invocation requires user approval. Used
+    here to exercise the full HTTP pipeline through ``ResponsesHostServer``
+    when the hosted agent is a ``WorkflowAgent`` containing a tool-approval
+    flow.
+    """
+
+    def __init__(
+        self,
+        name: str,
+        *,
+        tool_name: str = "delete_file",
+        tool_arguments: dict[str, Any] | None = None,
+        approval_request_ids: Sequence[str] | None = None,
+        final_text: str = "done",
+    ) -> None:
+        self.id = str(uuid.uuid4())
+        self.name = name
+        self.description: str | None = None
+        self._tool_name = tool_name
+        self._tool_arguments = tool_arguments or {"path": "/tmp/example"}
+        self._approval_request_ids: list[str] = list(approval_request_ids) if approval_request_ids else []
+        self._final_text = final_text
+        self.run_count = 0
+        self.last_run_messages: list[Message] = []
+
+    def create_session(self, **kwargs: Any) -> AgentSession:
+        return AgentSession()
+
+    def get_session(self, *, service_session_id: str, **kwargs: Any) -> AgentSession:
+        return AgentSession()
+
+    def _next_request_id(self) -> str:
+        # Stable across calls: when the workflow checkpoint round-trips through
+        # restore, ``AgentExecutor`` re-invokes the inner agent during replay.
+        # We must surface the *same* approval request id on each invocation so
+        # the workflow's pending-request id matches the id the test echoes
+        # back as ``mcp_approval_response``.
+        if self._approval_request_ids:
+            return self._approval_request_ids[0]
+        return str(uuid.uuid4())
+
+    def _build_approval_request(self) -> Content:
+        request_id = self._next_request_id()
+        function_call = Content.from_function_call(
+            call_id=request_id,
+            name=self._tool_name,
+            arguments=self._tool_arguments,
+            additional_properties={"server_label": "test_server"},
+        )
+        return Content.from_function_approval_request(id=request_id, function_call=function_call)
+
+    @overload
+    def run(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None = ...,
+        *,
+        stream: Literal[False] = ...,
+        session: AgentSession | None = ...,
+        **kwargs: Any,
+    ) -> Awaitable[AgentResponse[Any]]: ...
+
+    @overload
+    def run(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None = ...,
+        *,
+        stream: Literal[True],
+        session: AgentSession | None = ...,
+        **kwargs: Any,
+    ) -> ResponseStream[AgentResponseUpdate, AgentResponse[Any]]: ...
+
+    def run(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None = None,
+        *,
+        stream: bool = False,
+        session: AgentSession | None = None,
+        **kwargs: Any,
+    ) -> Awaitable[AgentResponse] | ResponseStream[AgentResponseUpdate, AgentResponse]:
+        if stream:
+            return self._run_stream(messages=messages, **kwargs)
+        return self._run(messages=messages, **kwargs)
+
+    @staticmethod
+    def _normalize(
+        messages: str | Content | Message | Sequence[str | Content | Message] | None,
+    ) -> list[Message]:
+        if messages is None:
+            return []
+        if isinstance(messages, str):
+            return [Message(role="user", contents=[Content.from_text(text=messages)])]
+        if isinstance(messages, Message):
+            return [messages]
+        if isinstance(messages, Content):
+            return [Message(role="user", contents=[messages])]
+        result: list[Message] = []
+        for item in messages:
+            if isinstance(item, Message):
+                result.append(item)
+            elif isinstance(item, Content):
+                result.append(Message(role="user", contents=[item]))
+            else:
+                result.append(Message(role="user", contents=[Content.from_text(text=item)]))
+        return result
+
+    @staticmethod
+    def _approval_responses_in(messages: list[Message]) -> list[Content]:
+        return [c for m in messages for c in m.contents if c.type == "function_approval_response"]
+
+    async def _run(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None = None,
+        **kwargs: Any,
+    ) -> AgentResponse:
+        normalized = self._normalize(messages)
+        self.last_run_messages = normalized
+        self.run_count += 1
+        if self._approval_responses_in(normalized):
+            return AgentResponse(messages=[Message("assistant", [Content.from_text(text=self._final_text)])])
+        approval = self._build_approval_request()
+        return AgentResponse(messages=[Message("assistant", [approval])])
+
+    def _run_stream(
+        self,
+        messages: str | Content | Message | Sequence[str | Content | Message] | None = None,
+        **kwargs: Any,
+    ) -> ResponseStream[AgentResponseUpdate, AgentResponse]:
+        normalized = self._normalize(messages)
+        self.last_run_messages = normalized
+        self.run_count += 1
+        approvals = self._approval_responses_in(normalized)
+
+        async def _iter() -> AsyncIterator[AgentResponseUpdate]:
+            if approvals:
+                yield AgentResponseUpdate(
+                    contents=[Content.from_text(text=self._final_text)],
+                    role="assistant",
+                    author_name=self.name,
+                )
+                return
+            yield AgentResponseUpdate(
+                contents=[self._build_approval_request()],
+                role="assistant",
+                author_name=self.name,
+            )
+
+        return ResponseStream(_iter(), finalizer=AgentResponse.from_updates)
+
+
+def _build_text_workflow_agent(text: str) -> WorkflowAgent:
+    """Build a minimal ``WorkflowAgent`` whose inner agent emits a fixed text."""
+
+    class _TextAgent(SupportsAgentRun):
+        def __init__(self, name: str, text: str) -> None:
+            self.id = str(uuid.uuid4())
+            self.name = name
+            self.description: str | None = None
+            self._text = text
+
+        def create_session(self, **kwargs: Any) -> AgentSession:
+            return AgentSession()
+
+        def get_session(self, *, service_session_id: str, **kwargs: Any) -> AgentSession:
+            return AgentSession()
+
+        @overload
+        def run(
+            self,
+            messages: Any = ...,
+            *,
+            stream: Literal[False] = ...,
+            session: AgentSession | None = ...,
+            **kwargs: Any,
+        ) -> Awaitable[AgentResponse[Any]]: ...
+
+        @overload
+        def run(
+            self,
+            messages: Any = ...,
+            *,
+            stream: Literal[True],
+            session: AgentSession | None = ...,
+            **kwargs: Any,
+        ) -> ResponseStream[AgentResponseUpdate, AgentResponse[Any]]: ...
+
+        def run(
+            self,
+            messages: Any = None,
+            *,
+            stream: bool = False,
+            session: AgentSession | None = None,
+            **kwargs: Any,
+        ) -> Awaitable[AgentResponse] | ResponseStream[AgentResponseUpdate, AgentResponse]:
+            text = self._text
+            name = self.name
+
+            async def _aresult() -> AgentResponse:
+                return AgentResponse(messages=[Message("assistant", [Content.from_text(text=text)])])
+
+            async def _aiter() -> AsyncIterator[AgentResponseUpdate]:
+                yield AgentResponseUpdate(
+                    contents=[Content.from_text(text=text)],
+                    role="assistant",
+                    author_name=name,
+                )
+
+            if stream:
+                return ResponseStream(_aiter(), finalizer=AgentResponse.from_updates)
+            return _aresult()
+
+    inner = _TextAgent("text-agent", text)
+
+    @executor
+    async def start(messages: list[Message], ctx: WorkflowContext[AgentExecutorRequest]) -> None:
+        await ctx.send_message(AgentExecutorRequest(messages=messages, should_respond=True))
+
+    workflow = WorkflowBuilder(start_executor=start).add_edge(start, inner).build()
+    return WorkflowAgent(workflow=workflow, name="Text Workflow Agent")
+
+
+def _build_approval_workflow_agent(
+    *,
+    approval_request_id: str,
+    tool_name: str = "delete_file",
+    tool_arguments: dict[str, Any] | None = None,
+    final_text: str = "done",
+) -> tuple[WorkflowAgent, _ToolApprovalWorkflowAgentMock]:
+    """Build a ``WorkflowAgent`` whose inner agent emits a tool approval request."""
+    mock_agent = _ToolApprovalWorkflowAgentMock(
+        name="approval-agent",
+        tool_name=tool_name,
+        tool_arguments=tool_arguments or {"path": "/tmp/secret.txt"},
+        approval_request_ids=[approval_request_id],
+        final_text=final_text,
+    )
+
+    @executor
+    async def start(messages: list[Message], ctx: WorkflowContext[AgentExecutorRequest]) -> None:
+        await ctx.send_message(AgentExecutorRequest(messages=messages, should_respond=True))
+
+    workflow = WorkflowBuilder(start_executor=start).add_edge(start, mock_agent).build()
+    workflow_agent = WorkflowAgent(workflow=workflow, name="Approval Workflow Agent")
+    return workflow_agent, mock_agent
+
+
+class TestWorkflowAgentHosting:
+    """End-to-end HTTP tests for ``ResponsesHostServer`` hosting a ``WorkflowAgent``.
+
+    These tests drive ``_handle_inner_workflow`` through the ASGI stack:
+    they exercise checkpoint write/restore (multi-turn) and the
+    tool-approval round-trip path, which is the primary differentiator
+    relative to the regular agent path.
+    """
+
+    async def test_basic_text_response(self) -> None:
+        workflow_agent = _build_text_workflow_agent("hello from workflow")
+        server = _make_server(workflow_agent)
+
+        resp = await _post(server, input_text="hi", stream=False)
+        assert resp.status_code == 200
+        body = resp.json()
+        assert body["status"] == "completed"
+
+        text_found = any(
+            part.get("type") == "output_text" and part.get("text") == "hello from workflow"
+            for item in body["output"]
+            if item["type"] == "message"
+            for part in item.get("content", [])
+        )
+        assert text_found, f"Expected workflow output text in {body['output']}"
+
+    async def test_basic_text_response_streaming(self) -> None:
+        workflow_agent = _build_text_workflow_agent("hello stream")
+        server = _make_server(workflow_agent)
+
+        resp = await _post(server, input_text="hi", stream=True)
+        assert resp.status_code == 200
+        events = _parse_sse_events(resp.text)
+        types = _sse_event_types(events)
+        assert types[0] == "response.created"
+        assert types[-1] == "response.completed"
+        assert "response.output_text.delta" in types
+        text_done = [e for e in events if e["event"] == "response.output_text.done"]
+        assert any(e["data"]["text"] == "hello stream" for e in text_done)
+
+    async def test_non_streaming_emits_mcp_approval_request_and_persists_to_storage(self) -> None:
+        workflow_agent, mock_agent = _build_approval_workflow_agent(approval_request_id="apr_wf_ns")
+        server = _make_server(workflow_agent)
+
+        resp = await _post(server, stream=False)
+        assert resp.status_code == 200
+        body = resp.json()
+        assert body["status"] == "completed"
+
+        approval_items = [it for it in body["output"] if it["type"] == "mcp_approval_request"]
+        assert len(approval_items) == 1
+        assert approval_items[0]["name"] == "delete_file"
+        assert approval_items[0]["server_label"] == "test_server"
+        approval_request_id = approval_items[0]["id"]
+
+        # The id surfaced over the wire is generated by the response stream
+        # builder; the original approval ``Content`` (carrying the inner
+        # ``function_call``) must be persisted under that id so the next
+        # turn can reconstruct it.
+        loaded = await server._approval_storage.load_approval_request(  # pyright: ignore[reportPrivateUsage]
+            approval_request_id
+        )
+        assert loaded.type == "function_approval_request"
+        assert loaded.function_call.name == "delete_file"  # type: ignore[attr-defined]
+        assert mock_agent.run_count == 1
+
+    async def test_streaming_emits_mcp_approval_request_and_persists_to_storage(self) -> None:
+        workflow_agent, mock_agent = _build_approval_workflow_agent(approval_request_id="apr_wf_st")
+        server = _make_server(workflow_agent)
+
+        resp = await _post(server, stream=True)
+        assert resp.status_code == 200
+
+        events = _parse_sse_events(resp.text)
+        types = _sse_event_types(events)
+        assert types[0] == "response.created"
+        assert types[-1] == "response.completed"
+
+        approval_request_id: str | None = None
+        for e in events:
+            if e["event"] != "response.output_item.added":
+                continue
+            item = e["data"].get("item") or {}
+            if item.get("type") == "mcp_approval_request":
+                approval_request_id = item.get("id")
+                break
+        assert approval_request_id is not None
+
+        loaded = await server._approval_storage.load_approval_request(  # pyright: ignore[reportPrivateUsage]
+            approval_request_id
+        )
+        assert loaded.type == "function_approval_request"
+        assert mock_agent.run_count == 1
+
+    async def test_round_trip_approval_response_resumes_workflow_agent(self) -> None:
+        """Two-turn HTTP round-trip:
+
+        Turn 1 emits ``mcp_approval_request`` and writes a workflow
+        checkpoint under the response id. Turn 2 sends the
+        ``mcp_approval_response`` with ``previous_response_id`` set, so the
+        host restores the checkpoint, the WorkflowAgent routes the
+        approval response back to the paused inner agent, and the inner
+        agent emits the final assistant text.
+        """
+        workflow_agent, mock_agent = _build_approval_workflow_agent(
+            approval_request_id="apr_wf_rt",
+            final_text="done with approval",
+        )
+        server = _make_server(workflow_agent)
+
+        first = await _post(server, stream=False)
+        assert first.status_code == 200
+        first_body = first.json()
+        first_response_id = first_body["id"]
+        approval_items = [it for it in first_body["output"] if it["type"] == "mcp_approval_request"]
+        assert len(approval_items) == 1
+        approval_request_id = approval_items[0]["id"]
+        assert mock_agent.run_count == 1
+
+        second_payload: dict[str, Any] = {
+            "model": "test-model",
+            "input": [
+                {
+                    "type": "mcp_approval_response",
+                    "approval_request_id": approval_request_id,
+                    "approve": True,
+                }
+            ],
+            "stream": False,
+            "previous_response_id": first_response_id,
+        }
+        second = await _post_json(server, second_payload)
+        assert second.status_code == 200
+        second_body = second.json()
+        assert second_body["status"] == "completed"
+
+        # The inner agent must have been resumed (restore replay + new turn).
+        # Restore call is a no-op for the mock (no input); the new-turn call
+        # delivers the approval response, so run_count grows by at least 1.
+        assert mock_agent.run_count >= 2
+
+        # The final assistant text from the resumed inner agent surfaces in
+        # the HTTP output.
+        text_pieces = [
+            part.get("text", "")
+            for item in second_body["output"]
+            if item["type"] == "message"
+            for part in item.get("content", [])
+            if part.get("type") == "output_text"
+        ]
+        assert any("done with approval" in t for t in text_pieces), (
+            f"expected resumed workflow output, got {second_body['output']}"
+        )
+
+        # The new-turn invocation of the inner agent must have received the
+        # approval response routed back through WorkflowAgent.
+        approval_responses = [
+            c for m in mock_agent.last_run_messages for c in m.contents if c.type == "function_approval_response"
+        ]
+        assert len(approval_responses) == 1
+        assert approval_responses[0].approved is True  # type: ignore[attr-defined]
+
+    async def test_round_trip_approval_response_streaming(self) -> None:
+        """Streaming variant of the round-trip: turn 2 is requested with
+        ``stream=true`` and surfaces the resumed text as SSE events."""
+        workflow_agent, mock_agent = _build_approval_workflow_agent(
+            approval_request_id="apr_wf_rt_st",
+            final_text="streamed-done",
+        )
+        server = _make_server(workflow_agent)
+
+        first = await _post(server, stream=False)
+        first_body = first.json()
+        first_response_id = first_body["id"]
+        approval_request_id = next(it["id"] for it in first_body["output"] if it["type"] == "mcp_approval_request")
+
+        second = await _post_json(
+            server,
+            {
+                "model": "test-model",
+                "input": [
+                    {
+                        "type": "mcp_approval_response",
+                        "approval_request_id": approval_request_id,
+                        "approve": True,
+                    }
+                ],
+                "stream": True,
+                "previous_response_id": first_response_id,
+            },
+        )
+        assert second.status_code == 200
+        events = _parse_sse_events(second.text)
+        types = _sse_event_types(events)
+        assert types[0] == "response.created"
+        assert types[-1] == "response.completed"
+
+        text_done = [e for e in events if e["event"] == "response.output_text.done"]
+        assert any("streamed-done" in e["data"]["text"] for e in text_done)
+        assert mock_agent.run_count >= 2
+
+    async def test_round_trip_approval_response_rejected(self) -> None:
+        """Sending ``approve=False`` must surface as ``approved=False`` to the
+        inner agent on resume."""
+        workflow_agent, mock_agent = _build_approval_workflow_agent(
+            approval_request_id="apr_wf_reject",
+            final_text="acknowledged",
+        )
+        server = _make_server(workflow_agent)
+
+        first = await _post(server, stream=False)
+        first_body = first.json()
+        first_response_id = first_body["id"]
+        approval_request_id = next(it["id"] for it in first_body["output"] if it["type"] == "mcp_approval_request")
+
+        second = await _post_json(
+            server,
+            {
+                "model": "test-model",
+                "input": [
+                    {
+                        "type": "mcp_approval_response",
+                        "approval_request_id": approval_request_id,
+                        "approve": False,
+                    }
+                ],
+                "stream": False,
+                "previous_response_id": first_response_id,
+            },
+        )
+        assert second.status_code == 200
+
+        approval_responses = [
+            c for m in mock_agent.last_run_messages for c in m.contents if c.type == "function_approval_response"
+        ]
+        assert len(approval_responses) == 1
+        assert approval_responses[0].approved is False  # type: ignore[attr-defined]
+
+
+# endregion

python/packages/gemini/README.md

@@ -10,6 +10,10 @@ pip install agent-framework-gemini --pre
 
 The Gemini integration enables Microsoft Agent Framework applications to call Google Gemini models with familiar chat abstractions, including streaming, tool/function calling, and structured output.
 
+## Structured Output
+
+Gemini structured output can be configured with either a Pydantic model in `response_format`, a JSON schema mapping in `response_format`, or a Gemini-specific `response_schema`. Declarative agents that define `outputSchema` pass that schema through `response_format`.
+
 ## Authentication
 
 The connector supports both `google-genai` authentication modes.

python/packages/gemini/agent_framework_gemini/_chat_client.py

@@ -109,8 +109,8 @@ class GeminiChatOptions(ChatOptions[ResponseModelT], Generic[ResponseModelT], to
             or ``types.Tool`` objects returned by ``get_code_interpreter_tool``, ``get_web_search_tool``,
             ``get_mcp_tool``, ``get_file_search_tool``, or ``get_maps_grounding_tool``.
         tool_choice: How the model picks a tool. One of ``'auto'``, ``'none'``, or ``'required'``.
-        response_format: Pydantic model type for structured JSON output. The response text is
-            parsed into the model and exposed via ``ChatResponse.value``.
+        response_format: Pydantic model type or JSON schema mapping for structured JSON output.
+            The response text is parsed and exposed via ``ChatResponse.value``.
         instructions: Extra system-level instructions prepended to the system message.
 
     Not supported, and passing these raises a type error:
@@ -255,6 +255,29 @@ _OPTION_CONSUMED_KEYS: frozenset[str] = frozenset({
 
 _OPTION_EXCLUDE_KEYS: frozenset[str] = _OPTION_EXPLICIT_KEYS | _OPTION_CONSUMED_KEYS
 
+_JSON_SCHEMA_TYPES: frozenset[str] = frozenset({
+    "array",
+    "boolean",
+    "integer",
+    "null",
+    "number",
+    "object",
+    "string",
+})
+
+_JSON_SCHEMA_KEYWORDS: frozenset[str] = frozenset({
+    "$defs",
+    "additionalProperties",
+    "allOf",
+    "anyOf",
+    "enum",
+    "items",
+    "oneOf",
+    "properties",
+    "required",
+    "type",
+})
+
 _FINISH_REASON_MAP: dict[str, FinishReasonLiteral] = {
     "STOP": "stop",
     "MAX_TOKENS": "length",
@@ -747,9 +770,13 @@ class RawGeminiChatClient(
                 continue
             kwargs[_OPTION_TRANSLATIONS.get(key, key)] = value
 
-        if options.get("response_format") or options.get("response_schema"):
+        response_format = options.get("response_format")
+        response_schema = options.get("response_schema")
+        if response_format is not None or response_schema is not None:
             kwargs["response_mime_type"] = "application/json"
-        if schema := options.get("response_schema"):
+        if response_schema is not None:
+            kwargs["response_schema"] = response_schema
+        elif (schema := self._extract_response_schema(response_format)) is not None:
             kwargs["response_schema"] = schema
         if tools := self._prepare_tools(options):
             kwargs["tools"] = tools
@@ -762,6 +789,48 @@ class RawGeminiChatClient(
 
         return types.GenerateContentConfig(**kwargs)
 
+    @staticmethod
+    def _extract_response_schema(response_format: Any) -> dict[str, Any] | None:
+        """Extract a Gemini response schema from supported mapping response_format shapes."""
+        if not isinstance(response_format, Mapping):
+            return None
+        mapping = cast("Mapping[str, Any]", response_format)
+
+        if (nested := RawGeminiChatClient._extract_response_schema(mapping.get("format"))) is not None:
+            return nested
+
+        json_schema = mapping.get("json_schema")
+        if isinstance(json_schema, Mapping):
+            schema = cast("Mapping[str, Any]", json_schema).get("schema")
+            if isinstance(schema, Mapping):
+                return dict(cast("Mapping[str, Any]", schema))
+
+        schema = mapping.get("schema")
+        if isinstance(schema, Mapping):
+            return dict(cast("Mapping[str, Any]", schema))
+
+        if RawGeminiChatClient._is_json_schema_mapping(mapping):
+            return dict(mapping)
+
+        return None
+
+    @staticmethod
+    def _is_json_schema_mapping(value: Mapping[str, Any]) -> bool:
+        """Return True when a mapping appears to be a JSON Schema rather than a response-format envelope."""
+        if not any(keyword in value for keyword in _JSON_SCHEMA_KEYWORDS):
+            return False
+
+        schema_type = value.get("type")
+        if schema_type is None:
+            return True
+        if isinstance(schema_type, str):
+            return schema_type in _JSON_SCHEMA_TYPES
+        if isinstance(schema_type, Sequence) and not isinstance(schema_type, (str, bytes)):
+            entries = cast("Sequence[object]", schema_type)
+            return all(isinstance(item, str) and item in _JSON_SCHEMA_TYPES for item in entries)
+
+        return False
+
     def _prepare_tools(self, options: Mapping[str, Any]) -> list[types.Tool] | None:
         """Translate the framework tool list into Gemini API tool objects.
 

python/packages/gemini/tests/test_gemini_client.py

@@ -9,7 +9,7 @@ from typing import Any
 from unittest.mock import AsyncMock, MagicMock, patch
 
 import pytest
-from agent_framework import Content, FunctionTool, Message
+from agent_framework import Agent, Content, FunctionTool, Message
 from google.genai import types
 from pydantic import BaseModel
 
@@ -915,6 +915,20 @@ async def test_response_format_populates_value_on_chat_response() -> None:
     assert response.value == Reply(text="hello")
 
 
+async def test_response_format_mapping_populates_value_on_chat_response() -> None:
+    """When response_format is a JSON schema mapping, ChatResponse.value must parse the response text."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text='{"text": "hello"}')]))
+    schema = {"type": "object", "properties": {"text": {"type": "string"}}}
+
+    response = await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": schema},
+    )
+
+    assert response.value == {"text": "hello"}
+
+
 async def test_response_schema_added_to_config() -> None:
     """Sets both response_mime_type and the raw schema on the config when response_schema is given."""
     client, mock = _make_gemini_client()
@@ -931,6 +945,284 @@ async def test_response_schema_added_to_config() -> None:
     assert config.response_schema == schema
 
 
+async def test_response_format_raw_json_schema_added_to_config() -> None:
+    """For declarative outputSchema, response_format may already be a raw JSON schema mapping."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text='{"answer": "hello"}')]))
+    schema = {
+        "type": "object",
+        "properties": {"answer": {"type": "string", "description": "The answer."}},
+        "required": ["answer"],
+    }
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": schema},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_mime_type == "application/json"
+    assert config.response_schema == schema
+
+
+async def test_agent_default_options_response_format_raw_schema_added_to_config() -> None:
+    """Agent default_options is the path used by declarative outputSchema."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text='{"answer": "hello"}')]))
+    schema = {"type": "object", "properties": {"answer": {"type": "string"}}, "required": ["answer"]}
+    agent = Agent(client=client, default_options={"response_format": schema})
+
+    await agent.run("Hi")
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_mime_type == "application/json"
+    assert config.response_schema == schema
+
+
+async def test_response_format_complex_raw_json_schema_preserved() -> None:
+    """Nested declarative schemas should be forwarded without losing shape or constraints."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text='{"answer": "ok"}')]))
+    schema = {
+        "type": "object",
+        "properties": {
+            "answer": {"type": "string"},
+            "citations": {
+                "type": "array",
+                "items": {
+                    "type": "object",
+                    "properties": {
+                        "source": {"type": "string"},
+                        "confidence": {"type": "number"},
+                    },
+                    "required": ["source"],
+                    "additionalProperties": False,
+                },
+            },
+        },
+        "required": ["answer"],
+        "additionalProperties": False,
+    }
+
+    await client.get_response(
+        messages=[
+            Message(
+                role="user",
+                contents=[
+                    Content.from_text(
+                        "Summarize a long document while preserving citation metadata.\n" + ("context\n" * 128)
+                    )
+                ],
+            )
+        ],
+        options={"response_format": schema},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_schema == schema
+
+
+async def test_response_format_json_schema_envelope_added_to_config() -> None:
+    """OpenAI-style json_schema envelopes should still provide Gemini with the inner schema."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text='{"answer": "hello"}')]))
+    schema = {"type": "object", "properties": {"answer": {"type": "string"}}}
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": {"type": "json_schema", "json_schema": {"name": "Answer", "schema": schema}}},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_mime_type == "application/json"
+    assert config.response_schema == schema
+
+
+async def test_response_format_format_envelope_added_to_config() -> None:
+    """Responses-style format envelopes should also provide Gemini with the nested schema."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text='{"answer": "hello"}')]))
+    schema = {"type": "object", "properties": {"answer": {"type": "string"}}}
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": {"format": {"type": "json_schema", "name": "Answer", "schema": schema}}},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_mime_type == "application/json"
+    assert config.response_schema == schema
+
+
+async def test_response_format_direct_schema_key_added_to_config() -> None:
+    """Provider-normalized mappings with a direct schema key should be accepted."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text='{"answer": "hello"}')]))
+    schema = {"type": "object", "properties": {"answer": {"type": "string"}}}
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": {"schema": schema}},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_mime_type == "application/json"
+    assert config.response_schema == schema
+
+
+async def test_response_format_json_schema_envelope_preserves_empty_schema() -> None:
+    """An explicitly empty JSON schema is still a schema and should not be dropped as falsy."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text="{}")]))
+    schema: dict[str, Any] = {}
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": {"type": "json_schema", "json_schema": {"name": "AnyJson", "schema": schema}}},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_schema == schema
+
+
+async def test_response_format_anyof_raw_schema_added_to_config() -> None:
+    """Raw schemas without a type should still be recognized when they use JSON Schema keywords."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text='"ok"')]))
+    schema = {"anyOf": [{"type": "string"}, {"type": "number"}]}
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": schema},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_schema == schema
+
+
+async def test_response_format_union_type_raw_schema_added_to_config() -> None:
+    """JSON Schema union type arrays should be treated as raw schemas."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text='{"answer": "hello"}')]))
+    schema = {"type": ["object", "null"], "properties": {"answer": {"type": "string"}}}
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": schema},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_schema == schema
+
+
+async def test_response_format_json_object_does_not_set_schema() -> None:
+    """A JSON-object response_format requests JSON output but is not itself a Gemini response schema."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text="{}")]))
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": {"type": "json_object"}},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_mime_type == "application/json"
+    assert config.response_schema is None
+
+
+async def test_response_format_json_schema_without_inner_schema_does_not_set_schema() -> None:
+    """A json_schema envelope without a schema should not be mistaken for a raw JSON schema."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text="{}")]))
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": {"type": "json_schema", "json_schema": {"name": "MissingSchema"}}},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_mime_type == "application/json"
+    assert config.response_schema is None
+
+
+async def test_response_schema_takes_precedence_over_response_format_schema() -> None:
+    """An explicit Gemini response_schema should win when both schema options are present."""
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text="{}")]))
+    response_format_schema = {"type": "object", "properties": {"name": {"type": "string"}}}
+    response_schema = {"type": "object", "properties": {"id": {"type": "integer"}}}
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": response_format_schema, "response_schema": response_schema},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_schema == response_schema
+
+
+async def test_response_format_raw_schema_kept_with_tools() -> None:
+    """Structured output must still reach Gemini when function tools are present."""
+
+    def calculator(expression: str) -> str:
+        """Evaluate a simple expression."""
+        return expression
+
+    tool = FunctionTool(name="calculator", func=calculator)
+    client, mock = _make_gemini_client()
+    mock.aio.models.generate_content = AsyncMock(return_value=_make_response([_make_part(text='{"answer": "4"}')]))
+    schema = {"type": "object", "properties": {"answer": {"type": "string"}}, "required": ["answer"]}
+
+    await client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("What is 2 + 2?")])],
+        options={"tools": [tool], "response_format": schema},
+    )
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content.call_args.kwargs["config"]
+    assert config.response_schema == schema
+    assert config.tools is not None
+    assert config.tools[0].function_declarations[0].name == "calculator"
+
+
+async def test_streaming_response_format_raw_schema_added_to_config() -> None:
+    """Streaming requests use the same config path and should also forward raw schema mappings."""
+    client, mock = _make_gemini_client()
+    chunks = [_make_response([_make_part(text='{"answer": "hello"}')], finish_reason="STOP")]
+    mock.aio.models.generate_content_stream = AsyncMock(return_value=_async_iter(chunks))
+    schema = {"type": "object", "properties": {"answer": {"type": "string"}}}
+
+    stream = client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": schema},
+        stream=True,
+    )
+    async for _ in stream:
+        pass
+
+    config: types.GenerateContentConfig = mock.aio.models.generate_content_stream.call_args.kwargs["config"]
+    assert config.response_mime_type == "application/json"
+    assert config.response_schema == schema
+
+
+async def test_streaming_response_format_mapping_populates_final_value() -> None:
+    """Streaming responses should preserve mapping response_format for final value parsing."""
+    client, mock = _make_gemini_client()
+    chunks = [_make_response([_make_part(text='{"answer": "hello"}')], finish_reason="STOP")]
+    mock.aio.models.generate_content_stream = AsyncMock(return_value=_async_iter(chunks))
+    schema = {"type": "object", "properties": {"answer": {"type": "string"}}}
+
+    stream = client.get_response(
+        messages=[Message(role="user", contents=[Content.from_text("Hi")])],
+        options={"response_format": schema},
+        stream=True,
+    )
+    async for _ in stream:
+        pass
+
+    final = await stream.get_final_response()
+    assert final.value == {"answer": "hello"}
+
+
 async def test_streaming_response_format_passed_to_build_response_stream() -> None:
     """Verifies that response_format is forwarded to _build_response_stream when streaming
     so that structured output parsing works correctly on the final assembled response.

python/samples/02-agents/mcp/README.md

@@ -13,9 +13,12 @@ The Model Context Protocol (MCP) is an open standard for connecting AI agents to
 | **Agent as MCP Server** | [`agent_as_mcp_server.py`](agent_as_mcp_server.py) | Shows how to expose an Agent Framework agent as an MCP server that other AI applications can connect to |
 | **API Key Authentication** | [`mcp_api_key_auth.py`](mcp_api_key_auth.py) | Demonstrates API key authentication with MCP servers using `header_provider`, runtime invocation kwargs, and a command-line API key argument |
 | **GitHub Integration with PAT** | [`mcp_github_pat.py`](mcp_github_pat.py) | Demonstrates connecting to GitHub's MCP server using Personal Access Token (PAT) authentication |
+| **Long-Running Task** | [`mcp_long_running_task.py`](mcp_long_running_task.py) | Demonstrates transparent SEP-2663 long-running task handling for MCP tools that advertise `taskSupport=required`. Self-spawns a stdio MCP child server |
 
 ## Prerequisites
 
+Most samples in this folder use OpenAI:
+
 - `OPENAI_API_KEY` environment variable
 - `OPENAI_CHAT_MODEL` environment variable
 
@@ -23,3 +26,8 @@ Run `mcp_api_key_auth.py` with the MCP API key as the first command-line argumen
 
 For `mcp_github_pat.py`:
 - `GITHUB_PAT` - Your GitHub Personal Access Token (create at https://github.com/settings/tokens)
+
+For `mcp_long_running_task.py` (uses Azure OpenAI via Entra-ID):
+- Run `az login` once
+- `AZURE_OPENAI_ENDPOINT` - your Azure OpenAI resource endpoint, e.g. `https://<resource>.openai.azure.com/`
+- `AZURE_OPENAI_CHAT_MODEL` (or `AZURE_OPENAI_MODEL`) - the deployment name (e.g. `gpt-4o-mini`)

python/samples/02-agents/mcp/mcp_github_pat.py

@@ -46,6 +46,8 @@ async def github_mcp_example() -> None:
     # The MCP tool manages the connection to the MCP server and makes its tools available
     # Set approval_mode="never_require" to allow the MCP tool to execute without approval
     client = OpenAIChatClient()
+    # Note that the tool created here will be executed remotely by OpenAI, not locally by
+    # your application.
     github_mcp_tool = client.get_mcp_tool(
         name="GitHub",
         url="https://api.githubcopilot.com/mcp/",

python/samples/02-agents/mcp/mcp_long_running_task.py

@@ -0,0 +1,181 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+"""
+MCP Long-Running Task (SEP-2663) Example
+
+Demonstrates that ``MCPStdioTool`` transparently drives the MCP long-running
+task lifecycle for tools that advertise ``execution.taskSupport == "required"``.
+The agent observes a single function-call result; the framework handles the
+``tools/call`` → ``tasks/get`` (polled) → ``tasks/result`` sequence in the
+background.
+
+Run it as a single file. The script doubles as both the client and the stdio
+MCP child server (the child branch is selected via ``--server``):
+
+    python mcp_long_running_task.py
+
+Requirements:
+- Azure CLI sign-in (``az login``) — used for Entra-ID auth against Azure OpenAI.
+- ``AZURE_OPENAI_ENDPOINT`` — your Azure OpenAI resource endpoint, e.g.
+  ``https://<resource>.openai.azure.com/``.
+- ``AZURE_OPENAI_CHAT_MODEL`` (or ``AZURE_OPENAI_MODEL``) — the deployment name,
+  e.g. ``gpt-4o-mini``.
+
+This sample uses the lower-level ``mcp.server.lowlevel.Server`` so it can:
+1. Advertise a tool with ``execution=ToolExecution(taskSupport="required")``.
+2. Enable the SDK's experimental task support for the ``tasks/*`` lifecycle.
+"""
+
+import asyncio
+import sys
+from datetime import timedelta
+from typing import Any
+
+from agent_framework import Agent, MCPStdioTool, MCPTaskOptions
+from agent_framework.openai import OpenAIChatClient
+from azure.identity import AzureCliCredential
+from dotenv import load_dotenv
+
+load_dotenv()
+
+
+# ---------------------------------------------------------------------------
+# MCP stdio server (child-process branch)
+# ---------------------------------------------------------------------------
+
+
+async def _run_server() -> None:
+    """Run a minimal stdio MCP server exposing one long-running tool."""
+    import mcp.types as types
+    from mcp.server.lowlevel import Server
+    from mcp.server.stdio import stdio_server
+
+    server: Server[Any, Any] = Server("mcp-long-running-task-demo")
+    # Auto-registers handlers for tasks/get, tasks/result, tasks/cancel, tasks/list
+    # backed by an in-memory store.
+    server.experimental.enable_tasks()
+
+    @server.list_tools()
+    async def _list_tools() -> list[types.Tool]:  # pyright: ignore[reportUnusedFunction]
+        return [
+            types.Tool(
+                name="slow_summary",
+                description=(
+                    "Produces a short summary of the supplied text after simulating several seconds of expensive work."
+                ),
+                inputSchema={
+                    "type": "object",
+                    "properties": {
+                        "text": {
+                            "type": "string",
+                            "description": "Text to summarize.",
+                        }
+                    },
+                    "required": ["text"],
+                },
+                # Advertise that this tool MUST be invoked via the task lifecycle.
+                execution=types.ToolExecution(taskSupport="required"),
+            )
+        ]
+
+    @server.call_tool()
+    async def _call_tool(name: str, arguments: dict[str, Any]) -> Any:  # pyright: ignore[reportUnusedFunction]
+        if name != "slow_summary":
+            raise ValueError(f"Unknown tool: {name}")
+
+        ctx = server.request_context
+
+        async def _work(task: Any) -> types.CallToolResult:
+            await task.update_status("Thinking...")
+            await asyncio.sleep(15.0)
+            text: str = (arguments.get("text") or "").strip()
+            words = text.split()
+            preview = " ".join(words[:6]) + ("..." if len(words) > 6 else "")
+            summary = (
+                f"Summarized {len(words)} word(s). First few words: '{preview}'."
+                if words
+                else "No input text was provided."
+            )
+            return types.CallToolResult(
+                content=[types.TextContent(type="text", text=summary)],
+                isError=False,
+            )
+
+        if not ctx.experimental.is_task:
+            # Client invoked the tool without task augmentation. Return a hard
+            # error so a misconfigured client surfaces the problem clearly.
+            return types.CallToolResult(
+                content=[
+                    types.TextContent(
+                        type="text",
+                        text="'slow_summary' must be invoked as a task.",
+                    )
+                ],
+                isError=True,
+            )
+
+        return await ctx.experimental.run_task(_work)
+
+    async with stdio_server() as (read_stream, write_stream):
+        await server.run(read_stream, write_stream, server.create_initialization_options())
+
+
+# ---------------------------------------------------------------------------
+# Agent client (default branch)
+# ---------------------------------------------------------------------------
+
+
+async def _run_client() -> None:
+    mcp_tool = MCPStdioTool(
+        name="LongRunningDemo",
+        description="Demo MCP server exposing a tool that advertises taskSupport=required.",
+        command=sys.executable,
+        args=[__file__, "--server"],
+        # Optional: cap individual tasks at two minutes. The server may apply its
+        # own default if this is omitted.
+        task_options=MCPTaskOptions(default_ttl=timedelta(minutes=2)),
+    )
+
+    async with Agent(
+        client=OpenAIChatClient(credential=AzureCliCredential()),
+        name="LROAgent",
+        instructions=(
+            "You are a helpful assistant. Use the slow_summary tool when the user "
+            "asks for a summary. Wait for the result and present it directly."
+        ),
+        tools=mcp_tool,
+    ) as agent:
+        prompt = (
+            "Please summarize the following text using your slow_summary tool: "
+            "'The Model Context Protocol lets language models talk to external "
+            "tools and resources through a small JSON-RPC surface.'"
+        )
+
+        print("=== run() ===")
+        print(f"User: {prompt}")
+        response = await agent.run(prompt)
+        print(f"Agent: {response.text}\n")
+
+        print("=== run(stream=True) ===")
+        print(f"User: {prompt}")
+        print("Agent: ", end="", flush=True)
+        async for update in agent.run(prompt, stream=True):
+            if update.text:
+                print(update.text, end="", flush=True)
+        print()
+
+
+# ---------------------------------------------------------------------------
+# Entry point
+# ---------------------------------------------------------------------------
+
+
+def main() -> None:
+    if len(sys.argv) > 1 and sys.argv[1] == "--server":
+        asyncio.run(_run_server())
+        return
+    asyncio.run(_run_client())
+
+
+if __name__ == "__main__":
+    main()

python/samples/03-workflows/agents/handoff_workflow_as_agent.py

@@ -134,15 +134,11 @@ def handle_response_and_requests(response: AgentResponse) -> dict[str, HandoffAg
         if message.text:
             print(f"- {message.author_name or message.role}: {message.text}")
         for content in message.contents:
-            if content.type == "function_call":
-                if isinstance(content.arguments, dict):
-                    request = WorkflowAgent.RequestInfoFunctionArgs.from_dict(content.arguments)
-                elif isinstance(content.arguments, str):
-                    request = WorkflowAgent.RequestInfoFunctionArgs.from_json(content.arguments)
-                else:
-                    raise ValueError("Invalid arguments type. Expecting a request info structure for this sample.")
-                if isinstance(request.data, HandoffAgentUserRequest):
-                    pending_requests[request.request_id] = request.data
+            if content.type == "function_call" and content.name == WorkflowAgent.REQUEST_INFO_FUNCTION_NAME:
+                request_function_args = WorkflowAgent.RequestInfoFunctionArgs.from_dict(content.arguments)  # type: ignore
+                request_id = request_function_args.request_id
+                request_event = request_function_args.request_event
+                pending_requests[request_id] = request_event.data
 
     return pending_requests
 

python/samples/03-workflows/agents/workflow_as_agent_human_in_the_loop.py

@@ -3,10 +3,8 @@
 import asyncio
 import os
 import sys
-from collections.abc import Mapping
 from dataclasses import dataclass
 from pathlib import Path
-from typing import Any
 
 from agent_framework.foundry import FoundryChatClient
 from azure.identity import AzureCliCredential
@@ -141,28 +139,14 @@ async def main() -> None:
     # Handle the human review if required.
     if human_review_function_call:
         # Parse the human review request arguments.
-        human_request_args = human_review_function_call.arguments
-        if isinstance(human_request_args, str):
-            request: WorkflowAgent.RequestInfoFunctionArgs = WorkflowAgent.RequestInfoFunctionArgs.from_json(
-                human_request_args
-            )
-        elif isinstance(human_request_args, Mapping):
-            request = WorkflowAgent.RequestInfoFunctionArgs.from_dict(dict(human_request_args))
-        else:
-            raise TypeError("Unexpected argument type for human review function call.")
-
-        request_payload: Any = request.data
+        human_request_args = WorkflowAgent.RequestInfoFunctionArgs.from_dict(human_review_function_call.arguments)  # type: ignore
+        request_payload = human_request_args.request_event.data
         if not isinstance(request_payload, HumanReviewRequest):
             raise ValueError("Human review request payload must be a HumanReviewRequest.")
-
-        agent_request = request_payload.agent_request
-        if agent_request is None:
-            raise ValueError("Human review request must include agent_request.")
-
-        request_id = agent_request.request_id
+        if not request_payload.agent_request:
+            raise ValueError("Human review request must contain an agent_request.")
         # Mock a human response approval for demonstration purposes.
-        human_response = ReviewResponse(request_id=request_id, feedback="", approved=True)
-
+        human_response = ReviewResponse(request_id=request_payload.agent_request.request_id, feedback="", approved=True)
         # Create the function call result object to send back to the agent.
         human_review_function_result = Content(
             "function_result",

python/samples/04-hosting/foundry-hosted-agents/responses/09_foundry_skills/provision_skills.py

@@ -28,6 +28,7 @@ import zipfile
 from pathlib import Path
 
 from azure.ai.projects.aio import AIProjectClient
+from azure.ai.projects.models import CreateSkillVersionFromFilesBody
 from azure.core.exceptions import ResourceNotFoundError
 from azure.identity.aio import DefaultAzureCredential
 from dotenv import load_dotenv
@@ -68,8 +69,13 @@ async def main() -> None:
             name = skill_md.parent.name
             print(f"Provisioning skill '{name}' from {skill_md.relative_to(SKILLS_DIR.parent)}...")
             await _delete_skill_if_exists(project, name)
-            imported = await project.beta.skills.create_from_package(_zip_skill_md(skill_md))
-            print(f"  Imported skill '{imported.name}' (id={imported.skill_id}, has_blob={imported.has_blob}).")
+            imported = await project.beta.skills.create_from_files(
+                name,
+                content=CreateSkillVersionFromFilesBody(
+                    files=[(f"{name}.zip", _zip_skill_md(skill_md), "application/zip")]
+                ),
+            )
+            print(f"  Imported skill '{imported.name}' (id={imported.skill_id}, version={imported.version}).")
 
         print("Verifying skills via project.beta.skills.list()...")
         listed = {skill.name: skill async for skill in project.beta.skills.list()}
@@ -79,8 +85,8 @@ async def main() -> None:
             if skill is None:
                 raise RuntimeError(f"Skill '{name}' was imported but is not present in the project listing.")
             print(
-                f"  OK '{skill.name}': id={skill.skill_id}, "
-                f"description={skill.description!r}, has_blob={skill.has_blob}"
+                f"  OK '{skill.name}': id={skill.id}, "
+                f"description={skill.description!r}, default_version={skill.default_version}"
             )
 
     print("Done.")