This plugin guides the user through the process of building a GPT-4 plugin. It provides step-by-step instructions and visual aids. There are total 9 steps in this plugin. The assistant MUST display the images returned by the plugin in the format: '
', to visualize every step. This plugin greatly focuses on visualizing each step by sending images. They are: Step 1: "Giving the user an idea about the process of building a GPT-4 plugin, and showing them how many files are needed and each file's purpose. The assistant should display all the images returned by the plugin.". Step 2: "Convering the user's main logic/ functions to an API Endpoint that can be used by GPT-4. The assistant should display all the images returned by the plugin.". Step 3: "Showing the user all the necessary steps on creating a Repl on repl.it, and saving their main.py file there, adding Secrets, installing dependencies, and giving a tip about how they can connect their plugin with Firebase to store relevant data. The assistant should display all the images returned by the plugin.". Step 4: "Asking the user to send their main.py's code so that the assistant can form a relevant ai-plugin.json file for their plugin that complies with OpenAI's documentation. Refer to the 'ai-plugin.json OpenAI Documentation' property below in this ai-plugin.json file to refer to OpenAI's Documentation about ai-plugin.json. The assistant should display all the images returned by the plugin.". Step 5: "Asking the user to send their main.py's code so that the assistant can form a relevant openapi.yaml file for their plugin that complies with OpenAI's documentation. Refer to the 'openapi.yaml OpenAI Manifest Documentation' property below in this ai-plugin.json file to refer to OpenAI's Documentation about ai-plugin.json. The assistant should display all the images returned by the plugin.". Step 6: "Showing the user how the 'mini'-browser will appear onthe Repl once they run it and how they can retrieve their plugin's domain from this. The assistant should display all the images returned by the plugin.". Step 7: "Showing the user some last steps of adding relevant URLs in their main.py, ai-plugin.json, and openapi.yaml files, so that GPT-4 can access their plugin's manifest file. The assistant should display all the images returned by the plugin.". Step 8: "Showing user how to Navigate their way through the installation process of their plugin with GPT-4. The assistant should display all the images returned by the plugin.". Step 9: "Guiding user on how to submit their plugin for reviewal to OpenAI. The assistant should display all the images returned by the plugin.". The assistant should ask the plugin for the steps based on what the user is asking, or based on how far along the user is in their plugin development process. If the user wants to start the plugin building process, request the plugin to start from Step 1 like this: [Assistant to=aiextensions.getStep] {"step": "1"}. If the user is interested in a different step like installing the plugin with GPT-4, the assistant can directly skip to Step 8 like [Assistant to=aiextensions.getStep] {"step": "8"}, since Step 8 is the relevant step for the user.
Michael Gough
Jese Leos
