Timeout issue when running Orchestration with multiple agents

[ShinjiCodeEVA]

Hi team 👋,

I’m experimenting with the new agent orchestration APIs in Semantic Kernel and ran into a problem where my orchestration never completes — it times out even though I expected a single pass through the agents.

---

Error

Unhandled exception. System.TimeoutException: Orchestration did not complete within the allowed duration (00:01:00).
at Microsoft.SemanticKernel.Agents.Orchestration.OrchestrationResult1.GetValueAsync(Nullable1 timeout, CancellationToken cancellationToken)
at Program.

$(String[] args) in C:\Users\Harvie\source\repos\DotnetSemanticKernel\AgentOrchestration\Program.cs:line 46
at Program.(String[] args)

---

Repro code

using Microsoft.SemanticKernel;
using Microsoft.SemanticKernel.Agents;
using Microsoft.SemanticKernel.Agents.Orchestration.Sequential;
using Microsoft.SemanticKernel.Agents.Runtime.InProcess;

# pragma warning disable SKEXP0110

var builder = Kernel.CreateBuilder()
    .AddOpenAIChatCompletion(
        modelId: "gpt-4.1-mini-2025-04-14",
        apiKey: "sk-proj-XXXXX"); // dummy key

Kernel kernel = builder.Build();

ChatCompletionAgent analystAgent = new ChatCompletionAgent {
    Name = "Analyst",
    Instructions = "You are a marketing analyst. Given a product description, identify:\n- Key features\n- Target audience\n- Unique selling points",
    Kernel = kernel,
};

ChatCompletionAgent writerAgent = new ChatCompletionAgent {
    Name = "Copywriter",
    Instructions = "You are a marketing copywriter. Given a block of text describing features, audience, and USPs, compose a compelling marketing copy (around 150 words).",
    Kernel = kernel,
};

ChatCompletionAgent editorAgent = new ChatCompletionAgent {
    Name = "Editor",
    Instructions = "You are an editor. Given the draft copy, polish grammar, clarity, and tone. Output the improved copy as a single text block.",
    Kernel = kernel,
};

SequentialOrchestration orchestration = new(analystAgent, writerAgent, editorAgent)
{
    ResponseCallback = ResponseCallback
};

InProcessRuntime runtime = new InProcessRuntime();
await runtime.StartAsync();

var result = await orchestration.InvokeAsync(
    "An eco-friendly stainless steel water bottle that keeps drinks cold for 24 hours",
    runtime);

// Collect results
string output = await result.GetValueAsync(TimeSpan.FromSeconds(60));
Console.WriteLine($"\nRESULT: {output}");

// Callback sample
ValueTask ResponseCallback(ChatMessageContent response)
{
    Console.WriteLine($"# {response.AuthorName}\n{response.Content}");
    return ValueTask.CompletedTask;
}

[Ingremio]

@ShinjiCodeEVA I ran into the same problem yesterday. As a workaround, try using AzureCliCredential instead of the API key when calling AddAzureOpenAIChatCompletion() on the Semantic Kernel Builder.

Interestingly, when using an API key in the SK builder for ChatCompletionService, it works fine, but when trying to use ChatCompletionAgent, it doesn't.

Test code

// Add references
using Azure.Identity;
using Microsoft.Extensions.DependencyInjection;
using Microsoft.Extensions.Logging;
using Microsoft.SemanticKernel;
using Microsoft.SemanticKernel.Agents;
using Microsoft.SemanticKernel.Agents.Orchestration.Sequential;
using Microsoft.SemanticKernel.Agents.Runtime.InProcess;
using Microsoft.SemanticKernel.ChatCompletion;

internal class Program
{
    private const string ServiceId = "aoai";

    public static async Task Main()
    {
        // Clear the console
        Console.Clear();

        // Get configuration settings
        var endpoint = Environment.GetEnvironmentVariable("AZURE_OPENAI_ENDPOINT");
        var apiKey = Environment.GetEnvironmentVariable("AZURE_OPENAI_API_KEY");
        var deployment = Environment.GetEnvironmentVariable("AZURE_OPENAI_CHAT_DEPLOYMENT_NAME");
        var apiVersion = Environment.GetEnvironmentVariable("AZURE_OPENAI_API_VERSION");

        if (string.IsNullOrWhiteSpace(endpoint) || string.IsNullOrWhiteSpace(apiKey) || string.IsNullOrWhiteSpace(deployment) || string.IsNullOrEmpty(apiVersion))
        {
            Console.Error.WriteLine("Brak wymaganych zmiennych środowiskowych: AZURE_OPENAI_ENDPOINT / AZURE_OPENAI_API_KEY / AZURE_OPENAI_DEPLOYMENT / AZURE_OPENAI_API_VERSION.");
            return;
        }

        var builder = Kernel.CreateBuilder();
        // Add logging for verification
        builder.Services.AddLogging(b => b.AddConsole().SetMinimumLevel(LogLevel.Debug));

        // Registering the Azure OpenAI connector
        builder.AddAzureOpenAIChatCompletion(
            deploymentName: deployment,
            endpoint: endpoint
            //,apiKey: apiKey //if with API key then ChatCompletionService works but ChatCompletionAgent does not
            ,new AzureCliCredential()
            ,serviceId: ServiceId);

        var kernel = builder.Build();

        // Checking the connection for ChatCompletion service
        var pingService = kernel.GetRequiredService<IChatCompletionService>(ServiceId);
        using var cts = new CancellationTokenSource(TimeSpan.FromSeconds(30));
        await pingService.GetChatMessageContentAsync("ping", cancellationToken: cts.Token);

        // Create a summarizer agent
        var summarizerAgent = new ChatCompletionAgent
        {
            Name = "SummarizerAgent",
            Description = "An agent that summarize the customer's feedback ",
            Instructions =
                """
                Summarize the customer's feedback in one short sentence. Keep it neutral and concise.
                Example output:
                App crashes during photo upload.
                User praises dark mode feature.
                """
            ,Kernel = kernel
            ,Arguments = new KernelArguments(new PromptExecutionSettings
            {
                ServiceId = ServiceId,
            })
        };

        // Create a classifier agent
        var classifierAgent = new ChatCompletionAgent
        {
            Name = "ClassifierAgent",
            Description = "An agent that classify the feedback",
            Instructions =
                """
                Classify the feedback as one of the following: Positive, Negative, or Feature request.
                """
            ,Kernel = kernel
            ,Arguments = new KernelArguments(new PromptExecutionSettings
            {
                ServiceId = ServiceId,
            })
        };

        // Create a recommmended action agent
        var actionAgent = new ChatCompletionAgent
        {
            Name = "ActionAgent",
            Description = "An agent that suggest the next action",
            Instructions =
                """
                Based on the summary and classification, suggest the next action in one short sentence.
                Example output:
                Escalate as a high-priority bug for the mobile team.
                Log as positive feedback to share with design and marketing.
                Log as enhancement request for product backlog.
                """
            ,Kernel = kernel
            ,Arguments = new KernelArguments(new PromptExecutionSettings
            {
                ServiceId = ServiceId
            })
        };

        ValueTask AgentResponseCallback(ChatMessageContent response)
        {
            Console.WriteLine($"# {response.AuthorName}\n{response.Content}");
            return ValueTask.CompletedTask;
        }

        // Create a sequential orchestration with a response callback to observe the output from each agent.
        var orchestration = new SequentialOrchestration(summarizerAgent, classifierAgent, actionAgent)
        {
            ResponseCallback = AgentResponseCallback
        };

        var runtime = new InProcessRuntime();
        await runtime.StartAsync();

        // Initialize the input task
        var task =
            """
            I tried updating my profile picture several times today, but the app kept freezing halfway through the process. 
            I had to restart it three times, and in the end, the picture still wouldn't upload. 
            It's really frustrating and makes the app feel unreliable.
            """;

        // Invoke the orchestration with a task and the runtime
        var orchestrationResult = await orchestration.InvokeAsync(
            input: task,
            runtime: runtime
        );

        // Wait for the results
        var value = await orchestrationResult.GetValueAsync(TimeSpan.FromSeconds(30));
        Console.WriteLine($"\n****** Task Input ******\n{task}");
        Console.WriteLine($"\n***** Final Result *****\n{value}");

        // Stop the runtime when idle
        await runtime.RunUntilIdleAsync();

        Console.ReadLine();
    }
}

[printf-hoven]

Same issue here also

SOLUTION = DON'T LEAVE Description empty

ChatCompletionAgent analystAgent = new ChatCompletionAgent {
Name = "Analyst",
Instructions = "You are a marketing analyst. Given a product description, identify:\n- Key features\n- Target audience\n- Unique selling points",
Description = "--don't leave empty--",
Kernel = kernel,
};