Inspect grading outputs and summaries for quality issues across directories.
npx skills add https://github.com/AKCqhzdy/dse-subject-grading --skill grading-qa
Please help me install this Agent Skill. Skill: grading-qa Source: https://github.com/AKCqhzdy/dse-subject-grading/tree/main/skills-v3/grading-qa Command: npx skills add https://github.com/AKCqhzdy/dse-subject-grading --skill grading-qa
Phase 8 grading QA provides a structured, repeatable process to validate the accuracy and completeness of grading outputs, reducing risk of inconsistent scoring or missing files.
Run the grading-qa process on the current grading batch to generate the qa_summary.md.
High-intent search queries and answers about installing and using this skill.
Grading quality assurance systematically validates score consistency by performing random spot-checks on students, assessing score distributions, and verifying output file completeness. It flags significant cross-student disparities to ensure reported results are accurate and reliable.
To validate grading output completeness, the QA process checks that final_scores.json and qa_summary.md exist and are up-to-date. It also verifies data coherence across the output/, rubric/, and extracted/ directories.
You can perform random spot-checks by running a QA process that selects 2–3 students and uses a sub-agent to re-validate their per-question marks against the grading rubrics, ensuring scoring accuracy and consistency.
Yes, you can assess score distribution anomalies for a yearly grading batch. The QA process calculates mean, median, and standard deviation to identify outliers, helping instructors trust the reported results.
The best way to generate a qa_summary.md is to run a structured grading QA process on your current batch. It automatically documents all quality findings, cross-student consistency checks, and any necessary remediation steps.
Your grading QA process flags significant cross-student disparities to enforce consistency requirements. It identifies quality issues where scoring outputs deviate significantly, reducing the risk of inconsistent grading across the dataset.
zonta99Validate ResearchPacks and Implementation Plans against quality rubrics to enforce gating decisions.
MassLab-SIICompute per-student grades from basic_info.txt files and generate CSV and summary outputs.
christopheryeoSearches Google Calendar by day, date, week, subject, attendee, or email with AND filters.
ZurybrImplement adaptive rejection sampling for log-concave distributions in R.
