Today to store the result I need to pass as argument to a bash function like

gpt "Tell me a joke" >> file.txt

Which is fine, but the idea is to make a shurtcut for that and include diferent places to store, such as json, txt, clipboard... That can also later be used to populate a history, maybe can be added the following flags

• --store="local" : to store the response into a txt, can be also clipboard value too
• --history: This one will create a json file with the history and if it exists, will read the history and save together..

Use the history is not yet a feature, maybe later..

There is no need for that, remove and leave the one on __main__.py

This was good to build locally, but since this project auto publish on docker.io, there is no need to build locally, if for some reason that is needed, just build as development ...

Also, update documentation to fix the image name (gpt4shell) ->

For now, there is no context to what the gpt will answer, create a way to add context for example, a parameter that reads from a file. So that the gpt can give answers according to the need og the user.

One of the best use cases for this, is instead of gpt answer a generic question, you can make it specialize in convert human readable text into terminal commands, givin' this context to the gpt!

The support is preatty good now, but we can Try to add more feature to the result, try look at the Rich Docs

If there is no internet connection or other unforeseen error happens it shoults a big message to the user that doesn't help! Just create a nice error message and if is an unknown error, point the user to leave a feedback at github!

Instead of using default, leave the user choose with a new tag, --set-temperature=0.1. Add a new default to 0.3. Validate values between 0 and 1.