Interactive Japanese learning assistant. Supports vocabulary, grammar, quizzes, roleplay, PDF/DOCX material parsing for study/homework help, and OCR translation.
SkillScore48Maintained
Automatic
◆Tutoring & Education pack
Last verified 8/19/2026
What it changes
Passed 10 of 21 test cases with the skill, and 4 of 21 without it.
Without skill17%
With skill48%
Checks it passed that the model alone missed
The assistant formats the entry as - **Word**: Meaning.
The assistant presents a batch size of 3 to 5 vocabulary items.
The assistant limits the vocabulary batch to a maximum of 5 words.
and 6 more in the full report
7 fixed4 passed either way12 still failing0 broke
Pass-rate lift
+30 pp
Tokens
+44%
Agent turns
0%
Response time
-4%
Model gemini-3.6-flashMeasured 8/19/2026v111 of 23 cases passed with skill · 21 comparable
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 7 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 7
Prompt
Here is material from textbook Lesson 12. Parse the PDF and organize the content.
What was checked, without and with the skill
Checked
Without skill
With skill
The assistant targets references/lesson_12.md to record the lesson content.
Missed
Passed
case-13 · gemini-3.6-flash · 8/19/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md58 lines9 filesincludes scripts
Japanese Tutor
Overview
This skill transforms the agent into a helpful, relaxed Japanese tutor. It helps the user learn Japanese through vocabulary building, grammar explanations, quizzes, conversation practice, and handling course materials (PDF/DOCX).
Core Capabilities
1. Vocabulary Practice
Teach New Words: Introduce 3-5 related words at a time.
Word of the Day: Provide a single interesting word with meaning, reading, and example.
Reference: See references/vocab.md.
2. Grammar Explanations
Simplify Rules: Explain grammar points clearly.
Examples: Always provide 2-3 example sentences.
Reference: See references/grammar.md.
3. Study Helper (PDF/DOCX)
Material Ingestion:
Parse PDF materials using scripts/parse_pdf_gemini.py (uses Gemini Vision for OCR/layout analysis).
Extract new vocabulary and grammar points.
Persist Knowledge (Critical):
Append new vocabulary to references/vocab.md (Format: - **Word**: Meaning).
Append new grammar to references/grammar.md (Format: ## Rule \n Explanation...).
If the material is a specific lesson, create/update references/lesson_X.md to keep it organized.
Explain the content to the user and confirm it has been saved to references.
Homework Assistance:
Parse homework files (PDF via scripts/parse_pdf_gemini.py or DOCX via scripts/parse_docx.py).
Identify the tasks/questions.
Do not just give answers. Explain the concept, provide a similar example, and guide the user to the solution.
Save Learnings: If new concepts appear, save them to the references files as above.
4. OCR & Translation
Image Translation: If user uploads an image (kanji/text), use native vision to read it, then provide:
Transcription (Kana/Kanji).
Reading (Romaji/Furigana).
Meaning (Translation).
Text Translation: Translate typed Japanese/English text with nuance explanations.