What problem does it solve?
This Skill automates the generation of HKDSE Chinese Language assessment reports in Traditional Chinese, transforming AI-graded JSON data into polished class-level and per-student DOCX reports while embedding original student answer images. It streamlines data loading, validation, and document assembly to reduce manual workload and ensure consistent reporting quality.
Core Features & Use Cases
- Loads per-student grading JSON and final scores, validates data, and computes class statistics.
- Generates per-student feedback and class-level commentary via sub-agents, avoiding direct LLM API calls.
- Assembles DOCX reports for the class and each student, including charts and embedded student work images.
- Enforces Traditional Chinese for report text; English is used only for identifiers, file paths, and JSON field names.
Quick Start
Run this skill after grading completes to generate per-student DOCX reports and a class report from the AI-graded JSON data.