feat: 🚀 😃 convert repo into a monorepo 🚀 😃 (#51)

* Merge main into turbo-monrepo

* feat: rename apps

* feat: remove supbase videos and serve all from project

* feat: move supabase to root

* feat: update README

* docs: update link

apps/dashboard/src/routes/api/completion/+server.js

@@ -0,0 +1,43 @@
+import { Configuration, OpenAIApi } from 'openai-edge';
+import { OpenAIStream, StreamingTextResponse } from 'ai';
+import { OPENAI_API_KEY } from '$env/static/private';
+
+// Create an OpenAI API client (that's edge friendly!)
+const openAIConfig = new Configuration({
+  apiKey: OPENAI_API_KEY
+});
+const openai = new OpenAIApi(openAIConfig);
+
+// Set the runtime to edge for best performance
+export const config = {
+  runtime: 'edge'
+};
+
+const instruction = {
+  outline: 'a structured lesson outline',
+  note: 'a sample lesson note',
+  activities: 'few interactive lesson activities to my lecture'
+};
+
+export async function POST({ request }) {
+  const { prompt } = await request.json();
+  const { courseTitle, lessonTitle, type } = JSON.parse(prompt);
+  const response = await openai.createChatCompletion({
+    model: 'gpt-3.5-turbo',
+    messages: [
+      {
+        role: 'system',
+        content: 'Act like an assistant of a lecturer'
+      },
+      {
+        role: 'user',
+        content: `Generate ${instruction[type]} given the course title is "${courseTitle}" and the title of the lesson is "${lessonTitle}". Format in HTML without any styling. MOST IMPORTANT DON'T include the title of the course and don't include the lesson title: "${lessonTitle}" in your output. Please make sure the content is well detailed.`
+      }
+    ],
+    stream: true
+  });
+  // Convert the response into a friendly text-stream
+  const stream = OpenAIStream(response);
+  // Respond with the stream
+  return new StreamingTextResponse(stream);
+}

apps/dashboard/src/routes/api/completion/customprompt/+server.js

@@ -0,0 +1,37 @@
+import { Configuration, OpenAIApi } from 'openai-edge';
+import { OpenAIStream, StreamingTextResponse } from 'ai';
+import { OPENAI_API_KEY } from '$env/static/private';
+
+// Create an OpenAI API client (that's edge friendly!)
+const openAIConfig = new Configuration({
+  apiKey: OPENAI_API_KEY
+});
+const openai = new OpenAIApi(openAIConfig);
+
+// Set the runtime to edge for best performance
+export const config = {
+  runtime: 'edge'
+};
+
+export async function POST({ request }) {
+  const { prompt } = await request.json();
+
+  const response = await openai.createChatCompletion({
+    model: 'gpt-3.5-turbo',
+    messages: [
+      {
+        role: 'system',
+        content: 'Act like an assistant of a lecturer or an educator'
+      },
+      {
+        role: 'user',
+        content: `${prompt}. DONT ADD ANY CSS STYLE TO THE HTML FORMATTING AND DON'T USE CODE HTML TAG`
+      }
+    ],
+    stream: true
+  });
+  // Convert the response into a friendly text-stream
+  const stream = OpenAIStream(response);
+  // Respond with the stream
+  return new StreamingTextResponse(stream);
+}

apps/dashboard/src/routes/api/completion/exerciseprompt/+server.js

@@ -0,0 +1,81 @@
+import { Configuration, OpenAIApi } from 'openai-edge';
+import { OpenAIStream, StreamingTextResponse } from 'ai';
+import { OPENAI_API_KEY } from '$env/static/private';
+
+// Create an OpenAI API client (that's edge friendly!)
+const openAIConfig = new Configuration({
+  apiKey: OPENAI_API_KEY
+});
+const openai = new OpenAIApi(openAIConfig);
+
+// Set the runtime to edge for best performance
+export const config = {
+  runtime: 'edge'
+};
+
+export async function POST({ request }) {
+  const { prompt } = await request.json();
+  const { questionNumber, optionNumber, lessonNote } = JSON.parse(prompt);
+
+  const refinedPrompt = `
+Please note this typescript interface:
+
+interface Question_type {
+  id: number;
+  label: any;
+}
+const QuestionTypes: Question_type[] = [
+  {
+    id: 1,
+    label: 'RADIO'
+  },
+  {
+    id: 2,
+    label: 'CHECKBOX'
+  },
+  {
+    id: 3,
+    label: 'TEXTAREA'
+  }
+];
+
+interface ExerciseTemplate {
+  title: string;
+  description: string;
+  questionnaire: {
+    questions: {
+      title: string;
+      name: string;
+      points: number; // a number out of 10 to score the student if student gets the correct answer
+      order: number; // a unique number from 0 to n+1, that orders the question in ascending order
+      question_type: Question_type;
+      options: {
+        label: string;
+        is_correct: boolean;
+      };
+    }[]
+  };
+}
+
+I am a teacher and I need your help evaluating my students after I have taught a lesson. Please generate a test with ${questionNumber} questions and ${optionNumber} options. The sourse of this test should come from my lesson note in thripple quotes: """${lessonNote}""". Format the quiz to match the ExerciseTemplate interface above and VERY IMPORTANTANTLY: ONLY RETURN JSON FORMAT AND DO NOT EXPLAIN ANTHING.
+NB: Ask a mixture of ${optionNumber} options and ${questionNumber} DIFFERENT Question Types(RADIO, CHECKBOX AND TEXTAREA), mark the correct answer.`;
+
+  const response = await openai.createChatCompletion({
+    model: 'gpt-3.5-turbo',
+    messages: [
+      {
+        role: 'system',
+        content: 'You only spit out JSON and nothing else, dont give any explanation.'
+      },
+      {
+        role: 'user',
+        content: refinedPrompt
+      }
+    ],
+    stream: true
+  });
+  // Convert the response into a friendly text-stream
+  const stream = OpenAIStream(response);
+  // Respond with the stream
+  return new StreamingTextResponse(stream);
+}