third-party-chat-history-storage

Storing Chat History in 3rd Party Storage

::: zone pivot="programming-language-csharp"

This tutorial shows how to store agent chat history in external storage by implementing a custom ChatMessageStore and using it with a ChatClientAgent.

By default, when using ChatClientAgent, chat history is stored either in memory in the AgentThread object or the underlying inference service, if the service supports it.

Where services do not require chat history to be stored in the service, it is possible to provide a custom store for persisting chat history instead of relying on the default in-memory behavior.

Prerequisites

For prerequisites, see the Create and run a simple agent step in this tutorial.

Install NuGet packages

To use Microsoft Agent Framework with Azure OpenAI, you need to install the following NuGet packages:

dotnet add package Azure.AI.OpenAI --prerelease
dotnet add package Azure.Identity
dotnet add package Microsoft.Agents.AI.OpenAI --prerelease

In addition, you'll use the in-memory vector store to store chat messages.

dotnet add package Microsoft.SemanticKernel.Connectors.InMemory --prerelease

Create a custom ChatMessage Store

To create a custom ChatMessageStore, you need to implement the abstract ChatMessageStore class and provide implementations for the required methods.

Message storage and retrieval methods

The most important methods to implement are:

• InvokingAsync - called at the start of agent invocation to retrieve messages from the store that should be provided as context.
• InvokedAsync - called at the end of agent invocation to add new messages to the store.

InvokingAsync should return the messages in ascending chronological order (oldest first). All messages returned by it will be used by the ChatClientAgent when making calls to the underlying xref:Microsoft.Extensions.AI.IChatClient. It's therefore important that this method considers the limits of the underlying model, and only returns as many messages as can be handled by the model.