This is a basic chatbot using the Ollama API, with the added bonus of Dark Clippy. It uses MS SemanticKernel to access an LLM running locally on Ollama. It's a work in progress (at least until Clippy takes over the world).
Clippy is courtesy of the Clippy.js repo (code originally by Smore.com, Cinnamon Software and, of course, Microsoft).
Meet Dark Clippy. The years have not been kind to the beloved Microsoft Office Assistant. He's been through some things. He's seen some things. He's done some things. Terrible things. He's not the same paperclip you remember.
Status:
If you haven't already, get Ollama and host an uncensored model like HammerAI/neuraldaredevil-abliterated on your local machine.
Adjust the appsettings.json file to point to your local Ollama server endpoint/port and preferred model.