Skip to content

Quick Start Guide

This guide walks you through setting up the Prisma Autograder GitHub Action from scratch in both execution modes.

Repo mode — step by step

1. Create the grading config directory

In your assignment repository template, create:

.github/
└── autograder/
    ├── criteria.json
    ├── feedback.json
    └── setup.json

2. Define your criteria

Create .github/autograder/criteria.json with your grading rubric:

{
  "test_library": "web_dev",
  "base": {
    "weight": 100,
    "subjects": [
      {
        "subject_name": "html_structure",
        "weight": 60,
        "tests": [
          {
            "file": "submission/index.html",
            "name": "has_tag",
            "parameters": [
              { "name": "tag", "value": "header" },
              { "name": "required_count", "value": 1 }
            ]
          }
        ]
      },
      {
        "subject_name": "css_styling",
        "weight": 40,
        "tests": [
          {
            "file": "submission/styles.css",
            "name": "check_flexbox_usage"
          }
        ]
      }
    ]
  }
}

3. Configure feedback (optional)

Create .github/autograder/feedback.json:

{
  "general": {
    "show_score": true,
    "show_passed_tests": false,
    "add_report_summary": true
  },
  "default": {}
}

4. Add the workflow

Create .github/workflows/classroom.yml:

name: Autograder
on:
  push:
    branches: [main]
  pull_request:
    branches: [main]
  workflow_dispatch:

jobs:
  grading:
    permissions: write-all
    runs-on: ubuntu-latest
    if: github.actor != 'github-classroom[bot]'
    steps:
      - name: Checkout repository
        uses: actions/checkout@v4
        with:
          path: submission

      - name: Run Autograder
        uses: webtech-network/autograder@main
        with:
          template-preset: "webdev"
          feedback-type: "default"
          include-feedback: "true"

Important

The path: submission in the checkout step is required. The Action expects all files under submission/.

5. Add secrets (if using AI feedback)

If you want AI-powered feedback:

      - name: Run Autograder
        uses: webtech-network/autograder@main
        with:
          template-preset: "webdev"
          feedback-type: "ai"
          include-feedback: "true"
          openai-key: ${{ secrets.ENGINE }}

Add your OpenAI API key as the repository secret ENGINE.

6. Verify

Push a commit and check:

  • ✅ The workflow runs successfully
  • ✅ The check run shows "Autograding Result" with the score
  • ✅ relatorio.md appears in the repo (if feedback enabled)

External mode — step by step

1. Set up the Autograder Cloud

Ensure your Cloud instance is running and accessible. Verify:

curl https://your-cloud-url/api/v1/health

2. Create a grading configuration

curl -X POST https://your-cloud-url/api/v1/configs \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "template_name": "input_output",
    "languages": ["python"],
    "include_feedback": true,
    "criteria_config": {
      "test_library": "input_output",
      "base": {
        "weight": 100,
        "subjects": [...]
      }
    },
    "feedback_config": { "general": { "show_score": true } },
    "setup_config": {}
  }'

Note the returned id (e.g. 42).

3. Add repository secrets

Secret Value
AUTOGRADER_CLOUD_URL https://your-cloud-url
AUTOGRADER_CLOUD_TOKEN Your integration token

4. Add the workflow

Create .github/workflows/autograder.yml:

name: Autograder (External)
on:
  push:
    branches: [main]
  workflow_dispatch:

jobs:
  grading:
    runs-on: ubuntu-latest
    if: github.actor != 'github-classroom[bot]'
    steps:
      - name: Checkout repository
        uses: actions/checkout@v4
        with:
          path: submission

      - name: Run Autograder
        uses: webtech-network/autograder@main
        with:
          execution-mode: "external"
          grading-config-id: "42"
          autograder-cloud-url: ${{ secrets.AUTOGRADER_CLOUD_URL }}
          autograder-cloud-token: ${{ secrets.AUTOGRADER_CLOUD_TOKEN }}
          submission-language: "python"
          feedback-type: "default"
          template-preset: "input_output"
          locale: "pt-br"

5. Verify

Push a submission and check:

  • ✅ The workflow completes without errors
  • ✅ A result appears on GET /api/v1/submissions/config/42

Demo repository

The webtech-network/demo-autograder repository is a complete working example of repo-mode grading.

What it demonstrates

  • A fully configured classroom.yml workflow
  • Real criteria.json for a web development assignment
  • Feedback configuration
  • Student submission files that get graded
  • Generated relatorio.md feedback artifact

Demo repository structure

demo-autograder/
├── .github/
│   ├── workflows/
│   │   └── classroom.yml          # Workflow definition
│   └── autograder/
│       ├── criteria.json           # Web dev grading rubric
│       ├── feedback.json           # Feedback display settings
│       └── setup.json              # Setup configuration
├── submission/
│   ├── index.html                  # Student HTML
│   ├── styles.css                  # Student CSS
│   └── app.js                      # Student JavaScript
└── relatorio.md                    # Generated feedback report

Demo workflow

name: Autograder
on:
  push:
    branches: [main]
  pull_request:
    branches: [main]
  workflow_dispatch:

jobs:
  grading:
    permissions: write-all
    runs-on: ubuntu-latest
    if: github.actor != 'github-classroom[bot]'
    steps:
      - name: Checkout repository
        uses: actions/checkout@v4
        with:
          path: submission

      - name: Run Autograder
        uses: webtech-network/autograder@main
        with:
          template-preset: "webdev"
          feedback-type: "default"
          include-feedback: "true"
          openai-key: ${{ secrets.ENGINE }}

Adapting for your course

  1. Fork/copy the demo structure into your assignment template
  2. Replace criteria.json with your rubric
  3. Adjust template-preset to match your assignment type:
    • "webdev" — HTML/CSS/JS assignments
    • "input_output" — stdin/stdout programs (Python, Java, C++)
    • "api" — REST API testing
  4. Keep the job named grading (required for check run export)
  5. Adjust triggers to match your workflow (push, PR, manual dispatch)

Common patterns

GitHub Classroom integration

For GitHub Classroom, the autograder integrates with the classroom bot:

if: github.actor != 'github-classroom[bot]'

This prevents the Action from running on the initial bot commit when creating student repos.

Multiple assignment types

You can have different workflows for different assignments in the same org by varying template-preset and criteria.json:

template-preset: "webdev"

template-preset: "input_output"

template-preset: "api"

Localization

Set the locale input to control feedback language:

locale: "pt-br"   # Portuguese (Brazil)
locale: "en"      # English (default)

See also