K-12 Exam Question Analysis

Problem: Single-Question Analysis Often Stops at the Answer

K-12 exam question analysis often stalls in two ways: it provides only the answer with a brief explanation, without curriculum-standard references or cognitive-level judgments; or it mixes multiple questions and question types, making option analysis, context decomposition, and diagram structure unstable. This skill narrows the task to one question at a time, turning the user's question text into saveable HTML analysis and diagram pages.

Workflow: Confirm Context First, Then Generate by Question Type

After triggering, the skill confirms grade band, subject, question number, and point value, then classifies the question as non-multiple-choice, single-choice, or themed multi-choice. Different types use different analysis templates: non-choice questions emphasize SOLO/cognitive levels, answer, reasoning chain, and study hints; choice questions require item-by-item option analysis, including distractor design patterns and common traps.

During analysis, it searches the relevant curriculum standards by subject, such as compulsory education or senior high school standards, and extracts content requirements, academic quality standards, and core literacy dimensions. If the question involves scientific frontiers, interdisciplinary contexts, or data charts, it may also supplement background information. The final output usually includes two files: questionX_analysis.html and questionX_diagram.html, presenting knowledge points, analysis structure, option comparison, and exam preparation tips.

Scope and Limits

It is suitable for teachers, curriculum reviewers, or students who need deep single-question breakdowns, but not for bulk grading, whole-paper scoring, or simple Q&A explanations. Users should provide the question text and be ready to confirm grade band, subject, and question number; if the user only asks how to solve it, the skill should answer directly instead of generating documents. Diagram output performs self-checks for SVG bounds, text overlap, content overflow, font hierarchy, color consistency, and curriculum-standard citations, but results still depend on template quality and the completeness of the supplied question text.