microsoft/semantic-kernel · critical · KeyNotFoundException

Could not find configuration section {caller}

Error message

Could not find configuration section {caller}

What it means

TestConfiguration.LoadRequiredSection uses IConfiguration.GetSection(caller).Get<T>() to bind a configuration section (named after the calling property via [CallerMemberName]) to a strongly-typed object. If the section is absent or binding yields null, it throws KeyNotFoundException. This enforces that required configuration sections (AzureOpenAIConfig, ApplicationInsightsConfig, etc.) exist at startup.

Source

Thrown at dotnet/samples/Demos/TelemetryWithAppInsights/TestConfiguration.cs:59

        }

        if (string.IsNullOrEmpty(caller))
        {
            throw new ArgumentNullException(nameof(caller));
        }

        return s_instance._configRoot.GetSection(caller).Get<T>();
    }

    private static T LoadRequiredSection<T>([CallerMemberName] string? caller = null)
    {
        var section = LoadSection<T>(caller);
        if (section is not null)
        {
            return section;
        }

        throw new KeyNotFoundException($"Could not find configuration section {caller}");
    }

#pragma warning disable CS8618 // Non-nullable field must contain a non-null value when exiting constructor.
    public class AzureOpenAIConfig
    {
        public string ChatDeploymentName { get; set; }
        public string ChatModelId { get; set; }
        public string Endpoint { get; set; }
        public string ApiKey { get; set; }
    }

    public class ApplicationInsightsConfig
    {
        public string ConnectionString { get; set; }
    }

    public class GoogleAIConfig
    {

View on GitHub (pinned to c028a0c7dc)

Solutions

1. Add the missing section to user secrets: 'dotnet user-secrets set "AzureOpenAIConfig:Endpoint" "<value>"' or the equivalent for your section.
2. Set the matching environment variables (use __ as the section separator, e.g., AzureOpenAIConfig__ApiKey).
3. Verify the [CallerMemberName] property name exactly matches the configuration section name you expect.
4. Ensure TestConfiguration's static constructor / initialization has run and the IConfigurationRoot is built from all expected sources.

Example fix

// before
private static T LoadRequiredSection<T>([CallerMemberName] string? caller = null)
{
    var section = LoadSection<T>(caller);
    if (section is not null) return section;
    throw new KeyNotFoundException($"Could not find configuration section {caller}");
}

// after — list all available sections for faster debugging
private static T LoadRequiredSection<T>([CallerMemberName] string? caller = null)
{
    var section = LoadSection<T>(caller);
    if (section is not null) return section;
    var available = string.Join(", ", s_instance._configRoot.GetChildren().Select(c => c.Key));
    throw new KeyNotFoundException(
        $"Could not find configuration section '{caller}'. Available sections: {available}");
}

// Check the section exists before binding
var section = config.GetSection(sectionName);
if (!section.Exists())
    throw new KeyNotFoundException($"Section '{sectionName}' missing. Available: {string.Join(", ", config.GetChildren().Select(c => c.Key))}");

bool SectionExists(IConfiguration config, string name) => config.GetSection(name).Exists();

When it happens

Trigger: Accessing a TestConfiguration property (e.g., TestConfiguration.AzureOpenAIConfig) when the corresponding configuration section is missing from user secrets, environment variables, or appsettings.json. The section name is derived from the property name via [CallerMemberName].

Common situations: User secrets not set for the required section; environment variable prefix doesn't match the section hierarchy; the section key name has a typo or casing mismatch; running tests in CI without configuring secrets; the IConfigurationRoot wasn't initialized properly (s_instance not built).

Related errors

• OpenAI/Azure OpenAI configuration was not found.
• Configuration is not provided.
• AI Chat Service type '{appConfig.RagConfig.AIChatService}' i
• AI Embedding Service type '{appConfig.RagConfig.AIEmbeddingS
• Vector store type '{appConfig.RagConfig.VectorStoreType}' is

AI-assisted analysis of microsoft/semantic-kernel@c028a0c7dc (2026-08-13). Data as JSON: /api/errors/bb6502734706fe88. Report an issue: GitHub.