skills.rest
AKCqhzdy
AKCqhzdyCOMMUNITY
Explore all skills by this author →
GitHub

hkdse-subject-grading

Extract PDF rubrics, calibrate reference data, and grade HKDSE exam papers.

Updated Aug 27, 2026
Search Tags:#calibration#rubrics#grading#pdf-extraction#hkdse#level-assignment
One-click install
npx skills add https://github.com/AKCqhzdy/dse-subject-grading --skill hkdse-subject-grading
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: hkdse-subject-grading
Source: https://github.com/AKCqhzdy/dse-subject-grading/tree/main/.github/skills/hkdse-subject-grading
Command: npx skills add https://github.com/AKCqhzdy/dse-subject-grading --skill hkdse-subject-grading

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Grades HKDSE elective subject exam papers by extracting rubric criteria from PDFs, calibrating against prior-year reference data, and providing per-student grading with per-question marks, total scores, and level assignment (1–5). It supports ICT, Music, Ethics and Religious Studies, Tourism and Hospitality Studies, Visual Arts, Biology, Economics, BAFS, and other subjects with typed written responses.

Core Features & Use Cases

  • Sub-agents are mandatory: Each student must be graded by a dedicated sub-agent.
  • Model-consistency and VLM rules: The main model and all sub-agents must be the same; VLM usage is restricted to allowed models and fallback flows.
  • Rubric calibration workflow: Extract rubrics from PDFs, calibrate against reference-year data, build empirical score-to-level mappings, and iteratively refine rubrics based on reference data.
  • End-to-end workflow: From rubric extraction to per-student grading, level assignment, and final score aggregation, with structured JSON outputs.

Quick Start

Provide the subject, year, and data paths to start the HKDSE grading workflow.

Frequently Asked Questions about hkdse-subject-grading

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate HKDSE grading for typed written responses?▼

Automate HKDSE grading by extracting rubrics from PDFs, calibrating against prior-year reference data, and producing per-student per-question scores and final level assignments. It supports subjects with typed written responses like ICT, Biology, and Economics.

How do I calibrate exam rubrics against prior-year reference data?▼

Calibrate exam rubrics by extracting criteria from PDFs and building empirical score-to-level mappings. The process refines rubrics iteratively based on reference-year data to ensure accurate per-student grading.

Can I grade multiple elective subjects using the same rubric-based scoring workflow?▼

Yes, you can grade multiple elective subjects using the same workflow. It supports subject-agnostic handling of typed written responses across ICT, Music, Tourism, Visual Arts, Biology, Economics, BAFS, and similar subjects.

Does HKDSE level assignment require dedicated sub-agents for each student?▼

HKDSE level assignment requires dedicated sub-agents for each student. The main model and all sub-agents must maintain strict model consistency, with controlled VLM fallback for robust rubric extraction.

What is the best way to extract rubrics from PDF exam papers for level assignment?▼

The best way to extract rubrics from PDFs is through a multi-phase workflow involving extraction, calibration, grading, and level division. This produces structured JSON outputs with precise score schemas to drive downstream reporting.

What limitations exist when using VLM fallback for rubric extraction?▼

VLM usage is restricted to allowed models and specific fallback flows only. It must maintain strict model consistency with the main model and all sub-agents, requiring robust rubric extraction and reference calibration to function properly.

Related Skills

Explore More Education & Research Skills→
juanrejuanre
3

structured-output

Extract structured data from LLM responses with JSON Schema validation.

Community
Intermediate
MassLab-SIIMassLab-SII
133

student-database

Compute per-student grades from basic_info.txt files and generate CSV and summary outputs.

Official
Basic
kthornkthorn
118

Building Paper Screening Rubrics

Develop and refine paper screening rubrics with versioned JSON criteria.

Community
Basic
inbharataiinbharatai
22

rubric-creator

Generate assessment rubrics with criteria, performance levels, and point values.

Community
Basic