Merge pull request #662 from westey-m/ichatclient-based-agent

Add docs for building an IChatClient based agent.

Author: westey-m

agent-framework/user-guide/agents/agent-types/TOC.yml

@@ -12,6 +12,8 @@
   href: openai-responses-agent.md
 - name: OpenAI Assistants Agents
   href: openai-assistants-agent.md
+- name: Agent based on any IChatClient
+  href: chat-client-agent.md
 - name: A2A Agents
   href: a2a-agent.md
 - name: Custom Agents

agent-framework/user-guide/agents/agent-types/chat-client-agent.md

@@ -0,0 +1,72 @@
+---
+title: Agent based on any IChatClient
+description: Learn how to use the Microsoft Agent Framework with any IChatClient implementation.
+zone_pivot_groups: programming-languages
+author: westey-m
+ms.topic: tutorial
+ms.author: westey
+ms.date: 09/25/2025
+ms.service: semantic-kernel
+---
+
+# Agent based on any IChatClient
+
+::: zone pivot="programming-language-csharp"
+
+The Microsoft Agent Framework supports creating agents for any inference service that provides a [`Microsoft.Extensions.AI.IChatClient`](/dotnet/ai/microsoft-extensions-ai#the-ichatclient-interface) implementation. This means that there is a very broad range of services that can be used to create agents, including open source models that can be run locally.
+
+In this document, we will use Ollama as an example.
+
+## Getting Started
+
+Add the required NuGet packages to your project.
+
+```powershell
+dotnet add package Microsoft.Extensions.AI.Agents
+```
+
+You will also need to add the package for the specific `IChatClient` implementation you want to use. In this example, we will use [OllamaSharp](https://www.nuget.org/packages/OllamaSharp/).
+
+```powershell
+dotnet add package OllamaSharp
+```
+
+## Creating a ChatClientAgent
+
+To create an agent based on the `IChatClient` interface, you can use the `ChatClientAgent` class.
+The `ChatClientAgent` class takes `IChatClient` as a constructor parameter.
+
+First, create an `OllamaApiClient` to access the Ollama service.
+
+```csharp
+using System;
+using Microsoft.Extensions.AI.Agents;
+using OllamaSharp;
+
+using OllamaApiClient chatClient = new(new Uri("http://localhost:11434"), "phi3");
+```
+
+The `OllamaApiClient` implements the `IChatClient` interface, so you can use it to create a `ChatClientAgent`.
+
+```csharp
+AIAgent agent = new ChatClientAgent(
+    chatClient,
+    instructions: "You are good at telling jokes.",
+    name: "Joker");
+```
+
+> [!IMPORTANT]
+> To ensure that you get the most out of your agent, make sure to choose a service and model that is well-suited for conversational tasks and supports function calling.
+
+::: zone-end
+::: zone pivot="programming-language-python"
+
+Documentation coming soon.
+
+::: zone-end
+
+## Using the Agent
+
+The agent is a standard `AIAgent` and supports all standard agent operations.
+
+See the [Agent getting started tutorials](../../../tutorials/overview.md) for more information on how to run and interact with agents.

agent-framework/user-guide/agents/agent-types/index.md

@@ -53,8 +53,7 @@ See the documentation for each service, for more information:
 |[OpenAI ChatCompletion](./openai-chat-completion-agent.md)|An agent that uses the OpenAI ChatCompletion service.|No|Yes|
 |[OpenAI Responses](./openai-responses-agent.md)|An agent that uses the OpenAI Responses service.|Yes|Yes|
 |[OpenAI Assistants](./openai-assistants-agent.md)|An agent that uses the OpenAI Assistants service.|Yes|No|
-|Ollama|An agent that uses the Ollama service.|Yes|No|
-|Any other ChatClient|You can also use any other ChatClient implementation to create a ChatClientAgent.|Varies|Varies|
+|[Any other ChatClient](./chat-client-agent.md)|You can also use any other [`Microsoft.Extensions.AI.IChatClient`](/dotnet/ai/microsoft-extensions-ai#the-ichatclient-interface) implementation to create an agent.|Varies|Varies|
 
 ## Complex custom agents