Extract structured text from exam PDFs and images into v1-schema JSON.
npx skills add https://github.com/AKCqhzdy/dse-subject-grading --skill extraction-v2
Please help me install this Agent Skill. Skill: extraction-v2 Source: https://github.com/AKCqhzdy/dse-subject-grading/tree/main/skills/extraction-v2 Command: npx skills add https://github.com/AKCqhzdy/dse-subject-grading --skill extraction-v2
Extract structured text from exam-paper PDFs / images (questions, student answers, rubrics) into v1-schema JSON (questions.json / rubrics.json) for downstream grading workflows.
Run the extraction-v2 pipeline on a PDF or directory to produce questions.json and rubrics.json and review the wrapped paper envelope artifacts in the outputs.
High-intent search queries and answers about installing and using this skill.
To extract structured data from exam PDFs, this pipeline reads questions, student answers, and rubrics, converting them into v1-schema JSON files (questions.json and rubrics.json) for downstream grading workflows. It processes both single and multi-PDF inputs.
Yes, you can extract questions and rubrics from images as well as PDFs. The pipeline supports both formats, embedding [IMAGE_DATA] tokens in the content when visuals matter and providing aligned images metadata to export page images later.
The extraction process handles diagrams by embedding [IMAGE_DATA] tokens directly into the structured text content when visuals matter. It also generates aligned images metadata, allowing you to export and reference the relevant page images later.
The v1-schema JSON format is a hierarchical structure that organizes extracted exam content into questions.json and rubrics.json. This standardized format ensures compatibility with downstream grading tools and legacy extraction workflows.
You configure multi-PDF extraction runs using YAML files. This allows you to define processing parameters for batch exam processing, while the pipeline maintains compatibility with legacy workflows via a symlinked scripts path and pluggable prompt module.
Yes, the exam extraction pipeline works with legacy grading systems by maintaining compatibility through a symlinked scripts path and a pluggable prompt module. It wraps results in a paper envelope for validation and downstream use.
bigbigboConvert DOCX exam papers into structured Markdown with linked images.
mietheScore composed artifacts against a blueprint rubric and output scores, issues, and pass.
sheng-jieSplit multi-page PDFs into individual pages or specified page ranges.
vinayakgExtract text and tables from PDFs into JSON or CSV.
