sGPT

A C# Program which using OpenAi API to help users to quickly communicate errors or any doubts encountered in their daily use of the computer through images in a short period of time.

s: simple / shot / silent

Why sGPT

Before the introduction of GPT-4o, users could interact with ChatGPT through voice mode but needed to rely on different models to process text, audio, and visual information separately. Imagine chatting with friends where one person is responsible for listening, another for seeing, and another for speaking. Such communication efficiency is clearly limited.

GPT-4o represents a significant improvement by integrating text, visual, and audio functions into a single model, greatly enhancing efficiency. When encountering problems in daily life, it can be challenging to describe issues using simple language.

For example, when a computer reports an error, shows a pop-up with an issue number, or a game or program fails to start, ordinary users might not understand these specialized error codes and situations. Searching for tutorials on the internet can be tedious and cumbersome.

With sGPT, users can send an image containing the error message to GPT with a single click. The program will automatically provide relevant instructions (customizable) to solve the user's problem step-by-step.

How to start

First, understand the structure of an OpenAI request and apply for an OpenAI API interface to use all functionalities.

Commonly used OpenAI API interfaces are divided into four categories:

• Dialogue
• Private Model Training
• General Purpose
• Image & Audio

Dialogue and Private Model Training are the most commonly used. The Dialogue class interface, the core for human-machine dialog, is subdivided into:

• Chat: Multi-round dialog.
• Completions: Single-round dialog, mainly for generating an article at one time, without the ability to interact in multiple dialogs.

Finding the right API endpoint

Refer to OpenAI's official API documentation. The endpoints will vary depending on the API functionality.

Create chat completion

Example request / default

curl https://api.openai.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-d '{
    "model": "gpt-4o",
    "messages": [
    {
        "role": "system",
        "content": "You are a helpful assistant."
    },
    {
        "role": "user",
        "content": "Hello!"
    }
    ]
}'

Example request / image input

curl https://api.openai.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-d '{
    "model": "gpt-4-turbo",
    "messages": [
    {
        "role": "user",
        "content": [
        {
            "type": "text",
            "text": "What'\''s in this image?"
        },
        {
            "type": "image_url",
            "image_url": {
            "url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"
            }
        }
        ]
    }
    ],
    "max_tokens": 300
}'

Example request / Streaming

curl https://api.openai.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-d '{
    "model": "gpt-4o",
    "messages": [
    {
        "role": "system",
        "content": "You are a helpful assistant."
    },
    {
        "role": "user",
        "content": "Hello!"
    }
    ],
    "stream": true
}'

---

Download all package that you need

dotnet add package MouseKeyHook --version 5.6.0
dotnet add package System.Net.Http
dotnet add package Newtonsoft.Json
dotnet add package Markdig
dotnet add package DotNetEnv
dotnet add package Microsoft.Web.WebView2 --version 1.0.2478.35 dotnet add package Microsoft.Web.WebView2

Create the Assets folder, if it doesn't already exist

New-Item -ItemType Directory -Force -Path .\Assets

Download Highlight.js JavaScript File

Invoke-WebRequest -Uri https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.3.1/highlight.min.js -OutFile .\Assets\highlight.min.js

Download Atom One Dark Style File

Invoke-WebRequest -Uri https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.3.1/styles/atom-one-dark.min.css -OutFile .\Assets\atom-one-dark.min.css

using list package to check

dotnet list package dotnet build netsh http add urlacl url=http://+:8080/ user=Everyone