Added support for passing conversation history as a JSON blob to prompt command

[richardhyy]

Implemented a new --conversation-json flag that allows users to supply a JSON-formatted conversation history when using the prompt command. This enhancement enables better interoperability.

Usage

Users can now pass conversation history in the form of a JSON string using the --conversation-json parameter.

Example command:

llm "What did you say?" -s "You are an AI assistant" --conversation-json "[{\"prompt\": \"Hi\", \"response\": \"Hello, how can I help you?\"}]"

Goal

The primary goal of this feature is to facilitate better integration between llm and other programs that handle conversation histories.

By allowing an external entity to manage the conversation state, this feature opens up new possibilities for developers to create more complex solutions which can leverage this CLI tool.

Still Needs

• [ ] Tests
• [ ] Documentation