docs: add developer guide for Reflection Widget

Author: Commanderk3

js/widgets/reflection-guide.md

@@ -0,0 +1,184 @@
+# Reflection Widget - Developer Guide
+
+## Overview
+
+The `ReflectionMatrix` class provides a chat-based interface for users to interact with AI mentors for reflecting on their music blocks project. It manages chat history, mentor selection, message exchange with a backend server, and report generation.
+
+Backend Code for the widget : [musicblocks_reflection_fastapi](https://github.com/Commanderk3/musicblocks_reflection_fastapi)
+
+## Key Features
+
+-   **Multi-mentor Chat:** Switch between AI mentors (Meta/Rohan, Music/Beethoven, Code/Steve).
+-   **Chat History:** Stores and renders conversation history.
+-   **Typing Indicator:** Shows animated "Thinking..." while awaiting responses.
+-   **Markdown Rendering:** Converts Markdown responses to HTML for display.
+-   **Analysis & Summary:** Fetches project analysis after sufficient chat history.
+-   **Local Storage:** Saves and retrieves analysis reports.
+-   **Export:** Download chat transcript as a text file.
+
+## Components
+
+-   **Widget Interface**: `js/widgets/reflection.js`
+-   **Block Definition**: `js/blocks/WidgetBlocks.js`
+-   **Registration**: `js/activity.js` (line 176) and `js/logo.js` (line 113)
+-   **CSS styling**: `css/activities.css` (line 2023 - 2072)
+
+## Methods
+
+-   **Widget Window Initialisation**
+
+    `init(activity)` : Triggers when the reflection block is clicked. It renders the widget and initializes the required data structures.
+
+---
+
+-   **Mentor Switching**
+
+    `changeMentor("code")` : Changes the active mentor and updates button highlights.
+
+---
+
+-   Sending Messages
+
+    `sendMessage()` : Pushes user message to history, displays it, and requests bot reply.
+
+---
+
+-   Bot Reply Handling
+
+    `botReplyDiv(message, user_query, md)` :
+
+    -   If `user_query` is true, sends message to backend and displays reply.
+    -   If `md` is true, renders reply as Markdown.
+
+---
+
+-   Analysis & Summary
+
+    `getAnalysis()` :
+
+    -   Requests project analysis from backend if chat history is sufficient.
+    -   Saves analysis to localStorage.
+
+---
+
+-   Local Storage
+
+    -   **Save:** `saveReport(data)` stores analysis.
+    -   **Read:** `readReport()` retrieves analysis.
+
+---
+
+-   Export Conversation
+
+    `downloadAsTxt(conversationData)` : Downloads chat history as a `.txt` file.
+
+---
+
+-   Markdown Rendering
+
+    `mdToHTML(md)` : Converts Markdown to HTML for display in chat.
+
+---
+
+## API Endpoints
+
+### 1. `/projectcode`
+
+-   **Method:** `POST`
+-   **Purpose:** Sends the exported project code to the backend to receive an algorithmic summary.
+-   **Payload:**
+    ```json
+    {
+        "code": "<project code string>"
+    }
+    ```
+-   **Response:**  
+    ```json
+    {
+        "algorithm": "<algorithm string>",
+        "message": "<first message>"
+    }
+    ```
+    or  
+    ```json
+    {
+        "error": "<error message>"
+    }
+    ```
+
+---
+
+### 2. `/chat`
+
+-   **Method:** `POST`
+-   **Purpose:** Sends a user query, chat history, mentor selection, and project algorithm to the backend to get an AI mentor's reply.
+-   **Payload:**
+    ```json
+    {
+      "query": "<user message>",
+      "messages": [ { "role": "...", "content": "..." }, ... ],
+      "mentor": "<mentor key>",
+      "algorithm": "<algorithm string>"
+    }
+    ```
+-   **Response:**  
+    `{ response: "<bot reply>" }`  
+    or  
+    `{ error: "<error message>" }`
+
+-   **Example `messages` structure**
+    ```json
+    [
+        {
+            "role": "meta",
+            "content": "hello"
+        },
+        {
+            "role": "user",
+            "content": "hi"
+        },
+        {
+            "role": "meta",
+            "content": "Hey there! I'm Rohan, the meta-mentor here. Tell me, what did you create in your project?"
+        }
+    ]
+    ```
+
+---
+
+### 3. `/analysis`
+
+-   **Method:** `POST`
+-   **Purpose:** Sends the chat history and summary to the backend to receive a project analysis report.
+-   **Payload:**
+    ```json
+    {
+      "messages": [ { "role": "...", "content": "..." }, ... ],
+      "summary": "<summary string>"
+    }
+    ```
+-   **Response:**  
+    `{ response: "<analysis report>" }`  
+    or  
+    `{ error: "<error message>" }`
+
+---
+
+**All endpoints expect and return JSON.**  
+**The backend is expected to run locally while development.**
+
+---
+
+### API testing
+
+```bash
+curl -X POST http://localhost:PORT/projectcode \
+     -H "Content-Type: application/json" \
+     -d '{
+           "code": "MUSICBLOCKS_PROJECT_CODE_HERE"
+         }'
+```
+
+**Note** : Save your Music Blocks project as an HTML file. Open the file to locate the embedded JSON code, and make sure to stringify the JSON before using it for testing.
+
+Written by : [Diwangshu Kakoty](https://github.com/Commanderk3)