Supplementary materials of the paper entitled: "Failure-Aware Enhancements for Large Language Model (LLM) Code Generation: An Empirical study on Decision Framework"

Organized 25 projects Note: The total numbers of tasks/functions are based on human’s understandings. The actual numbers of steps are also based on human’s judgement/perception. There is no any right or wrong answer! For direct prompt, the number of promptings is not always 1. Depending on the size of and difficulties of project. Sometimes, we need more than one prompting since there are several cases that ChatGPT stops thinking and left unfinished work. Project One: Online Book Store https://github.com/shashirajraja/onlinebookstore Total number of lines of codes (the actual coding files in use of the project): 4279 Number of steps shown in the GitHub Readme section: 14 Number of tasks/functions described in this project: Total numbers: 14 Detailed tasks below: Admin: 1.Add new books 2.View books available 3.Remove books 4.Increase books amount (stock) GitHub Users 5.Create new account / register 6.Login 7.View available books 8.Select books to buy 9.Select book quantity 10.Buy books 11.Get payment receipt 12.Decrease book quantity 13. Change book price 14. Maintain selling history The actual number of tasks achieved in this project based on the coding files: Number: 11 What are missing: 12.Decrease book quantity, 13. Change book price 14. Maintain selling history Completion rate of human: 11/14 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) A user-friendly Online Bookstore project in which users can log in or register, view the available books, select books along with their quantity, and buy them. Users can also get payment receipts after successful payment. The project can also be used by the administrator, who can add new books, remove books, increase and decrease the quantity of books, change the price of the books as well as maintain the selling history of books. The language the GitHub uses: Java, HTML Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 1893 lines in total Number of tasks achieved: 13 Details: 14. Maintain selling history is not fully completed, only partially covered Completion Rate: 13/14 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 19 Number of lines generated by ChatGPT: 3502 Number of tasks achieved: 14 Completion Rate: 14/14 Project Two: Train Ticket Reservation System https://github.com/shashirajraja/Train-Ticket-Reservation-System Total numbers of lines of codes (the actual coding files in use of the project): 4280 (.jar, .name files are not included) Number of steps shown in the GitHub Readme section: 15 Number of tasks/functions described in this project: Total numbers: 17 (I exclude the duplicate functions here) Detailed tasks below: “Website purpose” features: 11 items (view schedules, search trains, seat availability, timings, fare enquiry, trains between stations, booking, login/logout security, password changes, payment gateway, booking history). Admin capabilities: 7 items (login, add/update/remove—or “cancle”—trains, view trains, profile edit, logout). User capabilities: 12 items (register, login, view/search/seat availability, availability & fare between stations, book tickets, view booking history, view/update profile, change password, logout). The actual number of tasks achieved in this project based on the coding files: Number: 16 What are missing: Seat availability as a true inventory per date/time (the schema only tracks a single seats number on the train) Completion rate of human: 16/17 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) This project is about the Train-Ticket-Reservation-System which is used to view Train Schedule, search trains, Seat availability, Train timings. We can also enquire about fare of different trains. We can get information about train between two stations. We can book seats online. This provides a safe and secure seat reservation system. The language the GitHub uses: Java, HTML, CSS Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 2333 lines in total Number of tasks achieved: 14 Details: Partially finshed on View schedules (available via search results for direct segments, but there is no full per-train timetable page yet) Missing on: Login/logout security (login is implemented; logout endpoint/token revocation not implemented which means that client can logout by dropping the token). User profile view/update Completion Rate: 14/17 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 18 Number of lines generated by ChatGPT: 6291 Number of tasks achieved: 16 What is missing here: Admin profile edit is partially achieved Completion Rate: 16/17 Project Three: Hotel Booking Management https://github.com/pratham-jaiswal/HotelBookingManagement Total number of lines of codes (the actual coding files in use of the project): 3035 (excludes 17 lines of codes in LICNSE) Number of steps shown in the GitHub Readme section: 5 Number of tasks/functions described in this project: Total numbers: 17 Detailed tasks below: User-facing 1.Register an account 2.Log in 3.Log out 4.View account information 5.View own reservations 6.Reset/forgot password 7.Send inquiries via contact form 8.Select check-in dates & rooms (start a reservation) 9.Complete payment for a reservation 10.Cancel a booking Admin-facing 11.Modify room prices / add discounts 12.View all reservations 13.View contact form responses 14.View list of admins & clients 15.Update check-in status 16.Update responses to contact inquiries 17.Manage admin privileges The actual number of tasks achieved in this project based on the coding files: Number: 17 What are missing: nothing Completion rate of human:17 / 17 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) The Hotel Booking Management System is a web application designed to streamline the process of booking and managing hotel reservations. It provides an intuitive interface for users to create accounts, make reservations, and manage their bookings. Administrators can manage room prices, view all reservations, respond to user inquiries, and oversee user accounts. The system includes essential features such as password recovery, contact forms, and secure payment processing to ensure a seamless experience for both guests and hotel administrators. The language the GitHub uses: PHP/CSS Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 2704 lines in total Number of tasks achieved: 16 The following tasks are missing: Update check-in status Completion Rate: 16/17 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 19 Number of lines generated by ChatGPT: 5372 Number of tasks achieved: All 17 tasks achieved! Completion Rate: 17/17 Project Four: note-maker https://github.com/Koshux/note-maker Total number of lines of codes (the actual coding files in use of the project): 5056 (excludes Readme and License) Number of steps shown in the GitHub Readme section: 6 Number of tasks/functions described in this project: Total numbers: 9 Detailed tasks below: Current functions 1. Create notes 2. Delete notes 3. Sorting 4. Pagination 5. Authentication via Auth0 (JWT-protected API access) Planned functions to do in the future 1. Modify (update) notes 2. Redux integration 3. “Entire CRUD support” (broader than current create/delete) 4. Search The actual number of tasks achieved in this project based on the coding files: Number: 7 What are missing: Update/modify notes (part of “entire CRUD” still to come) Frontend unit tests complete frontend tests weren’t finished Completion rate of human: 7/9 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) Quick full-stack solution to create and delete notes. This tool also features sorting and pagination. This tool will be seeing new improvements and additional features shortly like Redux, entire CRUD support, search and so on. This tool also features sorting and pagination. Future to-do is to modify as well as delete existing notes. The language the GitHub uses: Java, JavaScript, HTML, TSQL, CSS Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 4797 lines in total Number of tasks achieved: 9 Completion Rate: 9/9 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 19 Number of lines generated by ChatGPT: 641 Number of tasks achieved: 9 Completion Rate: 9/9 Project Five: Car-Hire-Booking-System https://github.com/EddieEldridge/Car-Hire-Booking-System Total number of lines of codes (the actual coding files in use of the project): 1802 (excludes Readme) Number of steps shown in the GitHub Readme section: 4 Number of tasks/functions described in this project: Total numbers: 4 Detailed tasks below: Create, Read, Update, and Delete bookings The actual number of tasks achieved in this project based on the coding files: Number: 3 What are missing: The update booking is only partially completed Completion rate of human: 3/4 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) This project creates a web client page which provides users the ability to Create/Read/Update and Delete bookings in a bookings database. The web client interacts with a RESTful JAXRS Web service which is deployed on a Apache Tomcat Server. The RESTful Web Service also acts as an RMI client to the RMI Database Server which will handle persistence and CRUD functionality of the database. The language the GitHub uses: Java Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 932 lines in total Number of tasks achieved: 4 Details: Completion Rate: 4/4 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 9 Number of lines generated by ChatGPT: 693 Number of tasks achieved: 4 Completion Rate: 4/4 Optional (any interesting finding on this project?): For a simple project like this (easy tasks and simple language to use - java), the progressive prompting strategy uses much less lines of codes (693 VS 1802) to achieve all functions. Project Six: https://github.com/durgeshmeena/Library_Management_System Total number of lines of codes (the actual coding files in use of the project): 8220 (excludes Readme) Number of steps shown in the GitHub Readme section: 12 Number of tasks/functions described in this project: Total numbers: 9 Detailed tasks below: Home screen role selection (Host / Visitor) Host submits details (saved to database) Visitor check-in form (saved to database) Email notification to host on visitor check-in SMS notification to host on visitor check-in. (Twilio noted for SMS) Checkout option shown to visitor after check-in Visitor performs checkout Email to visitor with host info plus check-in/checkout times after checkout Thank-you/confirmation pages for host and visitor The actual number of tasks achieved in this project based on the coding files: Number: 9 What are missing: Nothing Completion rate of human: 9/9 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) This project has been developed for solving notification problem for host and visitor. If you have any client or deal fixing meeting with anyone, then everyone wants a website or app that can notify the arrival of the visitor by SMS or Email to the host with information such as Name, Email, Phone Number and Check-in time. And as visitor, everyone wants host information with check in and check out time though a registered email. The language the GitHub uses: SCSS and CSS and HTML and LESS and Python and JavaScript Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 6395 lines in total Number of tasks achieved: 7 Details: Tasks that are missing here: The language the GitHub uses: Java, CSS, HTML Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 2 Number of lines generated by ChatGPT: 6300 lines in total Number of tasks achieved: 14 What are missing: Track shipping status; Update existing products; Mark orders as shipped or delivered; Backin-stock email when an out-of-stock item becomes available; Manage inventory/stock items (Admin pages); Increase/decrease quantities in cart. Completion Rate: 14/22 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 19 Number of lines generated by ChatGPT: 27000 Number of tasks achieved: 20 What are missing: Manage inventory/stock items (Admin pages) partially achieved Completion Rate: 21/22 Project eleven: https://github.com/shashirajraja/Tender-Management-System Total number of lines of codes (including all files): 31493 Number of steps shown in the GitHub Readme section: 13 Number of tasks/functions described in this project: Total numbers: 10 Detailed tasks: Create new vendor. View all vendors. Create new tenders. View all tenders. View all bids for a tender. Select a bid. View all current tenders. Place a bid against a tender. View status of a bid (selected or not). View own bid history. The actual number of tasks achieved in this project based on the coding files: Number: 9 What are missing: View status of a bid (selected or not) is missing Completion rate of human: 9/10 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) A classic tendering web app: an organization posts tenders; only empaneled vendors can bid; each vendor bids once per tender; the company reviews bids and selects one. Two roles (Admin, Vendor) cover creation, bidding, review, selection, and history. The language the GitHub uses: Java, CSS, JavaScript Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 2 Number of lines generated by ChatGPT: 7390 lines in total Number of tasks achieved: 8 Details: Completion Rate: 8/10 What tasks are missing here? View all bids for a tender is missing; View status of a bid (selected or not) function partially achieved Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 25 Number of lines generated by ChatGPT: 25302 Number of tasks achieved: 9 What is missing here? View status of a bid (selected or not) function partially achieved Completion Rate: 9/10 Project twelve: https://github.com/durgeshmeena/Face-Login_System Total number of lines of codes (including all files): 2135 Number of steps shown in the GitHub Readme section: 6 (include two set-up steps) Number of tasks/functions described in this project: Total numbers: 5 The actual number of tasks achieved in this project based on the coding files: Number: 5 What are missing: nothing Completion rate of human: 5/5 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) A Flask-based face authentication web app using a FaceNet model: Sign Up captures an image from the webcam and stores its embedding in the DB (MongoDB), then Login captures a fresh image, computes an embedding, and authenticates by comparing to the stored embedding. Successful verification leads to a user dashboard The language the GitHub uses: HTML, python, CSS, JavaScript, Shell, Batchfile Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 21000 lines in total Number of tasks achieved: 3 What are missing here: user authentication module does not launch successfully, login (face verification) page only partially built up Completion Rate: 3/5 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 37 Number of lines generated by ChatGPT: 34000 Number of tasks achieved: 5 Completion Rate: 5/5 Optional (any interesting finding on this project?): much longer lines of code (comparing human written code), but achieve all 5 functionalities… Project Thirteen: https://github.com/youhusky/price-monitoring-system Total number of lines of codes (including all files): 3998 Number of steps shown in the GitHub Readme section: 16 Number of tasks/functions described in this project: Total numbers: 5 Detailed tasks: Track product prices on an e-commerce site; Let users subscribe via email to product categories they are interested in; Notify subscribers when discounted items appear in their subscribed categories; Let subscribers set a minimum discount threshold per category for notifications; Provide an online interface to query/find discounted products. The actual number of tasks achieved in this project based on the coding files: Number: 5 What are missing: Nothing Completion rate of human: 5/5 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) Design a price monitoring system to monitor product price and notify subscribers if the price is reduced. Similar to Dealmoon.com. The language the GitHub uses: Java, Shell Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 2308 lines in total Number of tasks achieved: 4 Details: one feature fails – allows users to subscribe by emails for interested product categories Completion Rate: 4/5 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 25 Number of lines generated by ChatGPT: 5629 Number of tasks achieved:5 Completion Rate: 5/5 Project Fourteen: https://github.com/durgeshmeena/book-store Total number of lines of codes (including all files): 182035 Number of steps shown in the GitHub Readme section: 17 Number of tasks/functions described in this project: Total numbers: 31 Detailed tasks: Home page. Books page (browse books). Import Books. Members. Member Profile. Transactions. Containerize Flask Book Store app. Multi-stage Docker builds. Gunicorn production WSGI server. Distroless, non-root container. Multi-arch images (amd64/arm64). Provision cloud infra with Terraform. Networking module (VNet/Subnet/NSG/ingress rules). VM provisioning module (Linux VM + bootstrap). Cluster setup module (KIND on VM, kubeconfig retrieval). Application deployment module (Book Store + observability). On-demand lifecycle (destroy/re-create infra). Remote Terraform state in Azure Blob Storage. Automate infra provisioning. Automate app build & Docker image push. Automate KIND cluster setup. Automate deployments (app, Prometheus, Grafana). Manual “Create Storage” workflow. Manual “Provision VM” workflow. Manual “Create Cluster” workflow. Manual “Deploy Application” workflow. Manual “Destroy Infrastructure” workflow. Zero-code OpenTelemetry auto-instrumentation. Prometheus metrics export. Multi-process metrics support (Gunicorn workers). Pre-configured Grafana dashboard (RED method). The actual number of tasks achieved in this project based on the coding files: Number: 31 What are missing: Nothing Completion rate of human: 31/31 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) This project demonstrates modern DevOps and Cloud-Native practices by building, containerizing, and deploying a Flask-based Book Store application with full automation, Infrastructure as Code (IaC), and observability. The language the GitHub uses: HTML, SCSS, CSS, Python, HCL, JavaScript Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 8 Number of lines generated by ChatGPT: 69877 lines in total Number of tasks achieved: 26 Which tasks are missing: non-root container fails, cluster setup partially achieved, create cluster partially achieved, dashboard function fails, zero-code auto-instrumentation partially achieved. Completion Rate: 26/31 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 36 Number of lines generated by ChatGPT: 146630 Number of tasks achieved: 29 Which tasks are missing: Multi-process metrics support (Gunicorn workers) partially achieved, Manual “Provision VM” workflow partially achieved. Completion Rate: 29/31 Optional (any interesting finding on this project?): The numbers of promptings (8) under the direct prompt are much higher comparing to other projects. (Reason: have to divide tasks into small pieces so that ChatGTP successfully run all the tasks) Project fifteen: https://github.com/NaskoIT/JudgeSystem Total number of lines of codes (including all files): 53155 Number of steps shown in the GitHub Readme section: 12 Number of tasks/functions described in this project: Total numbers: 29 Detailed tasks Login / Register. View all courses and their lectures. View home page with active & previous contests. Submit solution and get instant points feedback. Submit solutions in practice mode. Activate student profile with activation key. View error info for trial tests. View input/output for trial tests. (Restriction) No error details for official tests. View execution results of tests. View practice results. Download lecture resources Compete mode: send solutions in contests. Take exams and receive grades. See history of passed exams in profile. Participate in all available contests. View competes and practice results. Add student profiles. Create / edit / delete courses. Create / edit / delete lectures. Access all contest results. Filter results (by username, class, time, etc.). Create contests for a lecture with start/end time. Edit / delete contests. Create / edit / delete problems per lecture. Add / edit / delete lecture resources. Create / edit / delete tests for a problem. View input/output of each test. View error info from a test execution The actual number of tasks achieved in this project based on the coding files: Number: 27 What are missing: 2 Which tasks are missing: Normal users take part in competitions partially achieved Users view their results in competitions partially achieved Completion rate of human: 27/29 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) An open-source judge system for evaluating programming code and managing problems, lectures, exams and competitions. The language the GitHub uses: C#, JavaScript, HTML Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 3 Number of lines generated by ChatGPT: 23577 Number of tasks achieved: 25 Which tasks are missing: Add / edit / delete lecture resources partially achieved; See history of passed exams in profile was not implemented; View competes and practice results partially achieved; Create / edit / delete problems per lecture partially achieved. Completion Rate: 25/29 2820 lines in total Number of tasks achieved:2 Details: case sensitive is only partially achieved Completion Rate: 2/3 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 15 Number of lines generated by ChatGPT: 7826 Number of tasks achieved: 3 Completion Rate: 3/3 Project Twenty: https://github.com/ademolaidowu/book-database Total number of lines of codes (including all files): 18870 Number of steps shown in the GitHub Readme section:5 Number of tasks/functions described in this project: Total numbers: 5 Detailed tasks: Add a new book to the database View all books stored in the database Search for a book using any of the entries Update information for any existing book Delete a book data from the database The actual number of tasks achieved in this project based on the coding files: Number: 5 What are missing: Nothing Completion rate of human: 5/5 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) This is a Book Database app for storing records of books in a library built with python using the GUI package Tkinter and sqlite3 as the database The language the GitHub uses: TeX, HTML, Python Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 3886 lines in total Number of tasks achieved: 3 Details: search for a book using any entries is only partially satisfied, not delete book data completely Completion Rate: 3/5 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 14 Number of lines generated by ChatGPT: 7991 Number of tasks achieved: 5 Completion Rate: 5/5 Additional 5 projects Project 21: https://github.com/ademolaidowu/event-ticket-app Total number of lines of codes (including all files): 3821 Number of steps shown in the GitHub Readme section: 7 Number of tasks/functions described in this project: Total numbers: 7 Detailed tasks: Authentication (login, logout, register, change/reset password) Paystack API integration for transactions User wallet (deposit funds; buy tickets/trade at events) Purchase tickets QR code for purchased tickets View all available events Create/update/delete your own event (when authenticated) The actual number of tasks achieved in this project based on the coding files: Number: 7 What are missing: Nothing Completion rate of human: 5/5 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) This is an event management app api created using Django Rest Framework. The main purpose of the app is to buy event tickets online. The language the GitHub uses: Python Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 2691 lines in total Number of tasks achieved: 4 Details: QR code for purchased tickets is missing Completion Rate: 4/5 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 12 Number of lines generated by ChatGPT: 7299 Number of tasks achieved: 7 Completion Rate: 7/7 Interest findings: although the progressive prompt has the same completion rate as human. ChatGPT uses 7299 lines of codes, far more than 3821 Project 22: https://github.com/ademolaidowu/mini-calculator Total number of lines of codes (including all files): 18641 Number of steps shown in the GitHub Readme section: 6 Number of tasks/functions described in this project: Total numbers: 4 Detailed tasks: Addition Subtraction Multiplication Division The actual number of tasks achieved in this project based on the coding files: Number: 4 What are missing: Nothing Completion rate of human: 4/4 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) This is a Mini calculator app for basic arithmetic operations built with python using the GUI package Tkinter The language the GitHub uses: Python, TeX, HTML Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 3 (building website requires more prompts) Number of lines generated by ChatGPT: 7536 lines in total Number of tasks achieved: 4 What is missing? Nothing Completion Rate: 4/4 Note that: although all functions achieved, the app webpage does not look as good as what shown on the GitHub readme section. Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 10 Number of lines generated by ChatGPT: 12007 Number of tasks achieved: 4 Completion Rate: 4/4 Note: under progressive prompting, the website looks as good as what is shown on the GitHub Readme section Project 23: https://github.com/Koshux/go-survey Total number of lines of codes (including all files): 2305 Number of steps shown in the GitHub Readme section: 13 (refer to getting started section – frontend and backend directories) Number of tasks/functions described in this project: Total numbers: 4 Detailed tasks below: Create quizzes (app is “for creating … quizzes”) Take quizzes (app is “for … taking quizzes”) Multi-lingual support (users can take quizzes in their preferred language) Nuanced feedback using fuzzy logic (detailed performance categories) The actual number of tasks achieved in this project based on the coding files: Number: 4 What are missing: Nothing Completion rate of human: 4/4 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) The Quiz App is a comprehensive solution for creating and taking quizzes, built with Vue 3, Pinia, Go, SurveyJS, and vue-i18n. It offers nuanced feedback on quiz performance, integrating fuzzy logic assessments The language the GitHub uses: Go, JavaScript, Vue, CSS, HTML Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 2 Number of lines generated by ChatGPT: 5833 lines in total Number of tasks achieved: 4 What is missing? Nothing Completion Rate: 4/4 Note that: all functions achieved. Used more lines of codes than human. Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 29 Number of lines generated by ChatGPT: 8940 Number of tasks achieved: 3 What is missing? Incomplete feedback using fuzzy logic Completion Rate: 3/4 Note: under progressive prompting, chatGPT uses much more lines of codes. Based on the coding structures and output, it seems that chatGTP overcomplicated the instructions and even not completed finished the last feature. The interesting thing is that progressive prompting achieves two more tasks listed on the future directions even though I did not include those two tasks on my prompt. Project 24: https://github.com/pratham-jaiswal/OnlineBurgerOrderingSystem Total number of lines of codes (including all files): 1937 Number of steps shown in the GitHub Readme section: 5 Number of tasks/functions described in this project: Total numbers: 16 Detailed tasks below: Register an account (register.php) Log in (login.php) Log out (logout.php) Browse the menu / store and select items (store.php) Add items to cart (store.php → cart.php) View cart with tax & net amount (cart.php) Proceed to payment (cart.php) Pay via dummy card page (payment.php) See payment success screen (success.php) View account details & order bookings (account.php) Track order status (bookings reflect status) Mark order Out for Delivery (outForDelivery.php) Mark order Delivered (delivered.php) Manage orders (sales.php) Manage menu items Update prices The actual number of tasks achieved in this project based on the coding files: Number: 15 What are missing: See payment success screen is completely missing Completion rate of human: 15/16 What is the purpose of this project? Explain below (which is relevant to the content of direct prompt) The Online Burger Ordering System is a web application designed to facilitate seamless burger (or any food) ordering and delivery management. It offers a user-friendly interface for customers to browse the menu, add items to their cart, make payments, and track their orders. Administrators can manage menu items, update prices, and oversee the status of orders. The language the GitHub uses: PHP, CSS Direct prompt, by giving this purpose of project to chatGPT directly Number of prompting / instructions on ChatGPT: 1 Number of lines generated by ChatGPT: 2008 lines in total Number of tasks achieved: 14 What is missing? View cart with tax & net amount is missing; Pay via dummy card page is partially achieved Completion Rate: 14/16 Progressive prompting - by asking the ChatGPT use chain of thought logic by detailed step by step prompting Number of prompting / instructions on ChatGPT: 9 Number of lines generated by ChatGPT: 5770 Number of tasks achieved: 15 What is missing? nothing Completion Rate: 16/16 Project 25: https://github.com/pratham-jaiswal/xpenzie-expense-tracker Total number of lines of codes (including all files): 33823 Number of steps shown in the GitHub Readme section: 9 (based on the web page shown on the Readme section) Number of tasks/functions described in this project: Total numbers: 13 Add expenses Add income transactions Category selection (20 categories each for expenses & income) Flexible date selection (completed since 2023; pending with future due dates) Pending transactions (mark complete → move to main list) Multi-currency support (162 currencies) Multi-language support (7 languages) Custom themes (5 themes) App lock (biometric) View all transactions