[Enhancement]: Enable user to provide non-text reference information to the system (URLs, PDFs, images)

[MaxAnfilofyev]

Discussed in https://github.com/Pythagora-io/gpt-pilot/discussions/575

^{Originally posted by thelege2nd January 24, 2024} Hello, Please add a step before project initiation to either allow us to give new updates to the GPT pilot, for example, latest API documentation files like PDFs or perhaps it can pull the information from a given link to use it. It should be part of the questionnaire steps because it creates projects with outdated information and wrong practices.

[MaxAnfilofyev]

similar to

Discussed in https://github.com/Pythagora-io/gpt-pilot/discussions/441 Originally posted by garciaraul85 December 26, 2023 Can we add a feature to use GPT Vision API and create a UI/UX agent that takes blue-prints from the user to create UI/UX? I think that would be dope. Like you upload a sketch from a paper drawing to a professional blue-print of the UI whether its web, mobile or desktop and we let the agents code the UI/UX and finally we can add DALLE-2 or 3 or Midjourney create the logos and images for the UI/UX.

[blarvin]

#900 is related. Use file in project as source of truth and change...