Autograder - Finish-Up-A-Thon

Arthur Carvalho — Wed, 27 May 2026 22:54:36 +0000

This is a submission for the GitHub Finish-Up-A-Thon Challenge

This is: THE AUTOGRADER.

What I Built

As a child, I was not a good student, I was always disengaged and unfocused. I couldn't stay consistent on the schedule. After overcoming this, I built a strong commitment on helping people that struggle like i did. I built a strong commitment with education.

It's crazy how steve jobs nailed it when he talked about "connecting the dots". After working for 1 year as an english teacher and helping several people, i thought that was it, a good experience. However, in this year, when Rommel Carneiro my University teacher talked to me about working on a solution for student disengagement and teacher grading bottleneck, I could fully grasp the scenario. I had the whole perspective and was ready to contribute on narrowing down the solution.

The solution came to be the Autograder. It started from scratch, a single file with some lines of code. And we did something that was a success factor, at every new version of it, we had it be implemented in a learning process, which was everytime a bigger one. This allowed me to better map the key features to work on, based on student feedback.

We're far from a final product, but the current iteration is not that simple anymore. The Autograder is an extremely flexible, consistent and powerful tool for grading ANY kind of assignment. With support to AI evaluation and feedback generation, aswell as student code execution in isolated environments. We are proud to say that it has been used by more than 600 students now, being also applied in companies and in PUC Minas, my university.

Demo

webtech-network / autograder

Flexible, Consistent and Powerful Autograding Tool that grades and generates reports on your students submissions.

Autograder

An educational-standards-driven autograding tool that transforms assignment grading into an engaging learning experience.

Features • Architecture • Quick Start • Templates • Pipeline • API • GitHub Action • Docs • Community

Important

The Autograder is in active development. New features are being added continuously, and we welcome contributions from the community. We would love to hear your suggestions or feature requests! Don't hesitate on opening an issue on GitHub.

Overview

The Autograder is an advanced educational tool designed to efficiently and accurately grade student submissions using actual pedagogical standards. What makes it stand out is its highly elaborated grading methodology that follows teacher-configured rubrics and generates comprehensive, student-friendly feedback reports.

Why Autograder?

Teacher-Controlled Grading: Complete control over evaluation criteria with tree-structured rubrics
Educational Standards: Implements proper scoring categories (base, bonus, penalty) with weighted subjects
Multiple Assignment Types: Native support for Web Development, APIs, Command-Line Programs…

View on GitHub

How the grading pipeline works:

Complete documentation:

webtech-network.github.io

The Comeback Story

We implemented several AI analysis tests and stablished a flexible pipeline architecture. Our major focus were on resilience and extensibility. Adding new grading templates is now as easy as creating a python class.

We also created an entire learning platform called Prisma that runs on top of the autograder. It is being used in my univerisity by every freshman student in computer science. Teachers are being able to automatize validations for algorithm implementation, forbidden lib imports, forbidden keywors (like while loops or for loops), error handling and many more (check our full templates documentation).

My Experience with GitHub Copilot

Github copilot has been a major developer of this project. We have used it not only for PR reviews but for being one of the main developer that picks the issues we write down.

Members:

DEV Community: Arthur Carvalho