Enhancing AI Agents with Open API Integrations

March 30, 2025 by Anuraj

dotnet AI SemanticKernel

In this blog post, we’ll learn how to empower the AI agents with Tools. We will learn how to use Open API specification as a tool. For the demo, I created a Web API project - Story publishing API. Here is the Web API code.

var builder = WebApplication.CreateBuilder(args);

builder.Services.AddEndpointsApiExplorer();
builder.Services.AddSwaggerGen();

var app = builder.Build();

if (app.Environment.IsDevelopment())
{
    app.UseSwagger();
    app.UseSwaggerUI();
}

app.UseHttpsRedirection();

List<Story> stories = [];

app.MapGet("/stories", () =>
{
    return Results.Ok(stories);
})
.WithName("GetStories")
.WithDescription("Retrieve a list of all stories")
.WithOpenApi();

app.MapPost("/stories", (Story story) =>
{
    stories.Add(story);
    return Results.Created($"/stories/{story.GetHashCode()}",
        new { Message = "Story created successfully!" });
})
.WithName("CreateStory")
.WithDescription("Publish the story")
.WithOpenApi();

app.Run();

internal record Story(string Content, string Summary);

Run the application which will display the swagger UI. But we are more interested in the swagger.json file. We will be using the swagger.json file in our agent. Now in the Agent project we need to add reference of the Microsoft.SemanticKernel.Plugins.OpenApi nuget package. And we need to modify the code to import the Open API specification as plugin like this.

var kernel = serviceProvider.GetRequiredService<Kernel>();

kernel.Plugins.AddFromType<Plugins>();

await kernel.ImportPluginFromOpenApiAsync("StoryApi", 
    new Uri("https://localhost:7245/swagger/v1/swagger.json"));

var agent = new ChatCompletionAgent()
{
    Name = "StoryTellerAgent",
    Instructions = "You are a helpful AI Agent that can help to create small stories based on the user's topic " +
    "and display the story and the summary of the story. And publish the story",
    Kernel = kernel,
    Arguments = new KernelArguments(
        new PromptExecutionSettings()
        {
            FunctionChoiceBehavior = FunctionChoiceBehavior.Auto()
        })
};

In the above code, I modified the instructions again and included an instruction to publish the story - which will call the /stories POST endpoint. This way we can use Open API specification as tool / plugin for Semantic Kernel agents.

In the next post we will explore MCP - Model Context Protocol and how it can be used with Semantic Kernel.

Happy Programming

« Empowering AI Agents with Tools

Integrating Model Context Protocol Tools with Semantic Kernel »

Copyright © 2025 Anuraj. Blog content licensed under the Creative Commons CC BY 2.5 | Unless otherwise stated or granted, code samples licensed under the MIT license. This is a personal blog. The opinions expressed here represent my own and not those of my employer. Powered by Jekyll. Hosted with ❤ by GitHub