Math Error Tutor

Problem

Junior high math learning often stalls not because students cannot solve isolated problems, but because mistakes are not turned into tracked, re-practiced, and reviewable records. Oral explanations can miss step-by-step verification, precise knowledge-point mapping, and similar variants. Errors scattered in chat histories also do not become data for later exam paper generation or learning reports.

How It Works

math-learning-skills organizes the workflow around a math-error tutor role. Its core idea is to convert one mistake into a structured card, then use that card to drive practice, paper generation, and reporting:

• Error analysis: recognize the problem and geometry information from an image, detect multiple questions, and ask the student to select one. Diagnose it as what went wrong, why it went wrong, and how to avoid it, with step-by-step calculation and back-substitution checks.
• Adaptive practice: generate similar, variant, or comprehensive questions around the same knowledge point. It starts with three questions by default and can adjust difficulty after consecutive correct or incorrect answers; wrong submissions are written back to the error library.
• Smart exam generation: build a paper from selected knowledge points, difficulty, question types, and optional original items. Each answer is solved independently and checked for scale before the paper is exported from an HTML template and synchronized to IMA notes.
• Reports and synchronization: aggregate error cards, practice records, and reports, then archive them by subject area, knowledge point, and month in the IMA “Math Codex”. Images are left as placeholders, and geometry figures are described in text instead of SVG.

Boundaries

This skill targets Shanghai-edition math for grades 6 to 9 and focuses on error diagnosis, variant practice, and learning review. Non-math content is redirected to the main topic, and out-of-scope content is explicitly flagged. IMA synchronization depends on account authorization, an existing knowledge base, and the required folders; users must insert images manually in the notes, so it is not suitable for fully offline or note-less environments.