Get started by building a barebones chatbot

This is an example of a barebones chatbot that just waits for the user to say something, and then uses an LLM to generate and send a response.

The purpose is to illustrate the very basics of a workflow with OmniChain.

Note: Configured to use Ollama with the llama3.1:8b-instruct-q8_0 model.

Explanation

The Start node activates a CheckForNextMessage node which loops in on itself until it detects a new message.

Once a message from the user is detected, a BlockChat node is used to block OmniChain's integrated chat UI and a Response node is triggered.

The Response node requests a message to be built by a BuildMessage node, which uses an OllamaChatCompletion node to generate a response via an LLM hosted locally with Ollama.

The Ollama node receives the message history using a ReadSessionMessages node with a limit set to -1, which means the entire history is read.

Once the response is generated and sent to the user, another BlockChat node is used to unblock the chat UI and the CheckForNextMessage node is triggered to start waiting again.