HK

hkdse-subject-grading

Automates the grading of HKDSE elective subject exam papers using official rubrics and calibrated reference data.

Install

mkdir -p .claude/skills/hkdse-subject-grading && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/14912" && unzip -o skill.zip -d .claude/skills/hkdse-subject-grading && rm skill.zip

Installs to .claude/skills/hkdse-subject-grading

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Grades HKDSE elective subject exam papers for a specified subject. Use this when asked to grade typed student answer PDFs against an official rubric. Handles rubric extraction from PDFs (including split-part PDFs), reference calibration from prior-year labeled data, per-student grading with per-question mark allocation, total-score compilation, and level division (1–5). Subject-agnostic: works for ICT, Music, Religion, Tourism, Visual Arts, Biology, Economics, BAFS, and similar subjects with typed written responses.
521 chars✓ has a “when” triggerlonger than Claude Code's old 250-char listing cap (fine on current versions)
Advanced

Key capabilities

  • Grade HKDSE elective subject exam papers
  • Extract rubrics from PDF documents
  • Calibrate grading against prior-year labeled data
  • Perform per-question mark allocation for student answers
  • Compile total scores and assign levels (1-5)
  • Support subjects with typed written responses (e.g., ICT, Music, Biology)

How it works

The skill grades exam papers by extracting rubrics, calibrating against reference data, and then grading each student's answer PDF on a per-question basis using dedicated sub-agents and specific VLM models for text extraction.

Inputs & outputs

You give it
Typed student answer PDFs and official rubric PDFs
You get back
Per-student JSON results, compiled final scores, and level assignments (1-5)

When to use hkdse-subject-grading

  • Grading student exams
  • Extracting rubrics from PDFs
  • Assigning HKDSE levels
  • Compiling scores from exam parts

About hkdse-subject-grading

Processes student answer PDFs and compares them against extracted rubrics. It performs per-question grading, calculates total scores, and assigns levels (1-5) for various elective subjects.

>

When not to use it

  • For batch grading of multiple students in a single sub-agent call
  • For generating feedback or analysis text using LLM/VLM API calls
  • For subjects requiring dedicated Chinese handwriting scoring dimensions

Limitations

  • Sub-agents are mandatory for each student
  • VLM tasks are restricted to Kimi K2.5 ONLY
  • No LLM/VLM API calls for generating feedback or analysis text

How it compares

This approach uses a unified skill for various elective subjects, enforcing critical rules like mandatory sub-agents per student and specific VLM model usage, unlike general grading systems without such strict protocols.

Compared to similar skills

hkdse-subject-grading side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
hkdse-subject-grading (this skill)04moReviewAdvanced
scientific-critical-thinking187moReviewAdvanced
scholar-evaluation42moReviewIntermediate
tooluniverse-drug-research32moNo flagsAdvanced

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

scientific-critical-thinking

davila7

Evaluate research rigor. Assess methodology, experimental design, statistical validity, biases, confounding, evidence quality (GRADE, Cochrane ROB), for critical analysis of scientific claims.

1888

scholar-evaluation

K-Dense-AI

Systematically evaluate scholarly work using the ScholarEval framework, providing structured assessment across research quality dimensions including problem formulation, methodology, analysis, and writing with quantitative scoring and actionable feedback.

438

tooluniverse-drug-research

mims-harvard

Generates comprehensive drug research reports with compound disambiguation, evidence grading, and mandatory completeness sections. Covers identity, chemistry, pharmacology, targets, clinical trials, safety, pharmacogenomics, and ADMET properties. Use when users ask about drugs, medications, therapeutics, or need drug profiling, safety assessment, or clinical development research.

323

exam-question-pipeline

Leonamin

Convert professor-provided lecture PDFs, transcripts, prior exams, and generated drafts into Exam-Lab question-bank JSON with local-only material handling, source separation, exam-session metadata, validation, and app upload steps. Use when preparing midterm/final expected questions, converting cour

00

matlab-symbolic-math

matlab

Generate correct MATLAB code using the Symbolic Math Toolbox. Use when the user asks for symbolic computations, analytical solutions, symbolic differentiation/integration, equation solving, or converting symbolic results to numeric MATLAB functions. Also use when converting differential equations to

00

executing-marketing-campaigns

davila7

Plans, creates, and optimizes marketing campaigns including content strategy, social media, email, and analytics. Helps develop go-to-market strategies, campaign messaging, and performance measurement.

1181

Search skills

Search the agent skills registry