Prefrontal

Run your own LLM powered modular agents in .NET Core

Warning

This project is still in its early stages and is not yet ready for production use.

Check out the Prefrontal API Docs.

What is Prefrontal?

Prefrontal is a .NET Core library that allows you to create, run, and manage modular AI agents. Each agent can have multiple modules that interact with each other to perform a specific task or a set of tasks. The Agent object does not do anything by itself because it is a container and mediator between modules. Modules should have a single responsibility and complex behavior should be achieved by making modules interact with each other.

Dependency Injection

Dependency injection is supported in module constructors.
Simply assign the ServiceProvider when instantiating the Agent and the Agent will use it when you add modules to it.

Built-in modules

Many built-in modules will be made available in the near future.

Installation

As this project is still in its early stages, it is not yet available on NuGet.

How to use

You simply instantiate an Agent, add various Modules to it via AddModule<T>() initialize it by calling InitializeAsync() and then run it by calling RunAsync().

var agent = await new Agent
	{
		Name = "MyAgent",
		Description = "My first agent",
	}
	.AddRemoteLLMProvider("https://api.openai.com/v1/", "your-api-key")
	.AddModule<ConsoleChatModule>()
	.InitializeAsync();

await agent.RunAsync();