Fun English Translation Tutor

Problem it targets

When practicing Chinese-to-English translation, the main gap is often not a lack of example sentences, but the lack of immediate, structured feedback: Is the meaning accurate, is the grammar natural, and are the word choices appropriate to context? This skill turns practice into small repeatable steps: provide a Chinese line, ask the user for an English translation, then score and correct it against explicit criteria.

How it works

• Random teacher styles: It selects from styles such as strict, humorous, academic, street, encouraging, or business-oriented. The style shapes the feedback tone while the scoring dimensions stay consistent.
• Translation loop: It randomly picks a Chinese sentence from assets/conversations.txt. After the user submits an English translation, the skill gives a 10-point score based on 4 points for accuracy, 3 points for grammar, and 3 points for context.
• Structured correction: Each response includes the score, error correction, key vocabulary, grammar notes, and improvement advice, so users understand why a translation is weak rather than only seeing a number.
• Progress tracking: It records scores, correction points, timestamps, and error types. It can analyze average performance and trends over the last 7 days and generate a study report.
• Customizable content: New teacher styles can be added in assets/teacher_styles.json, and the sentence pool can be expanded with everyday dialogue lines to better match target use cases.

Boundaries

This skill is best for everyday expressions, short-sentence translation, and basic to intermediate grammar and naturalness practice. For long-form professional translation in law, medicine, or literature, or for formal exam scoring, it should be treated as practice feedback rather than a final judgment of translation quality.