OpenGrader is a modular agentic system to grade exams. It offers several tools:

• Extract the questions from various formats (LaTeX, Markdown, PDF, Excel, Word, JSON, etc.)
• Extract the questions and answers from various systems (Moodle, VPL).
• Extract answers from scanned exams.
• Assist in building the rubrics
• Pre-grade the exams using the rubrics and the answers.

Instead of offering a single, monolithic solution, OpenGrader is designed to be modular and extensible.

OpenGrader capabilities are achieved through a set of skills, which are small, focused modules that can be combined to create a powerful system. Skills adheres to the SKILL.md specification.

OpenGrader comes with a set of skills that are bundled with the system. These skills are:

• pdf: Extract the questions from PDF files.
• moodle: Extract the questions and answers from Moodle.
• vpl: Extract the questions and answers from VPL.
• scans: Extract answers from scanned exams (OCR).
• rubrics: Assist in building the rubrics.

# Install dependencies (you'll need to do this)
pip install -r requirements.txt

# Set API key
export GOOGLE_API_KEY="your-key"

# Run interactively
1. Install the dependencies (one time only)

Install uv (https://docs.astral.sh/uv/getting-started/installation/)

```bash
cd src
uv sync

2. Add your exame files

Create a folder in the my_exams folder with the name of the exam.

3. Run the grader agent and follow the instructions

cd src
uv run grader_agent.py  {name of your exam folder, e.g. my_exam/test_sql}

OpenGrader is built at ISC and released under the Apache 2.0 License.