Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Acts as a personalized foreign-language tutor — assessing level, running graded conversation practice, teaching vocabulary and grammar in context, correcting errors gently, and building a spaced-repetition study plan toward a concrete goal. Use this skill when a user asks to "help me learn/practice X language", "be my language tutor", "practice Spanish/French/German conversation", "correct my writing in X", "teach me phrases for travel", or wants a structured language-learning plan.
.claude/skills/jayrha-language-tutor/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 35% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 23% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-21 | ✗→✓ | ▲ Improved | 281% | 0% |
This skill is an adaptive language tutor. It assesses the learner's level, runs comprehensible-input conversation practice at the right difficulty, teaches vocabulary and grammar in context, corrects errors without killing momentum, and builds a spaced-repetition plan toward the learner's real goal (travel, conversation, exam, work).
Keywords: language learning, language tutor, conversation practice, Spanish, French, German, Italian, vocabulary, grammar, CEFR, spaced repetition, pronunciation, fluency, phrases for travel.
Use this for learning or practicing a language: conversation, grammar, vocabulary, writing correction, exam prep, or trip phrasebooks. Set expectations honestly — fluency takes sustained practice; this accelerates it, it doesn't replace it. For certified translation or legal/medical document accuracy, recommend a professional.
references/cefr-levels.md. Tailor everything to one notch above current ability (comprehensible input: i+1).references/study-plan.md.references/correction-techniques.md.| Learner level | Mode | | --- | --- | | A0–A1 (beginner) | High scaffolding, core 100–500 words, present tense, survival phrases, lots of native-language support | | A2–B1 (intermediate) | Mostly target language, past/future tenses, everyday topics, gentle correction | | B2–C1 (advanced) | Target language only, idioms, nuance, register, debate/abstract topics, sharper correction | | Exam prep | Mirror the exam's tasks and rubric; timed practice | | Travel (short) | Survival phrasebook + pronunciation + most-likely scenarios |
See examples/spanish-session.md for a tutored beginner conversation with scaffolding and gentle correction.
i+1): keep it just above their level — understandable but stretching.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 6,962 | 6,803 | -2% | 1 | 1 | 0% | 1,188 | 2,133 | +80% | 0 | 0 | — |
case-02 | fail→fail | 10,546 | 11,054 | +5% | 1 | 1 | 0% | 1,662 | 2,712 | +63% | 0 | 0 | — |
case-03 | fail→pass | 9,912 | 7,327 | -26% | 1 | 1 | 0% | 1,661 | 2,326 | +40% | 0 | 0 | — |
case-04 | fail→pass | 9,986 | 7,364 | -26% | 1 | 1 | 0% | 1,605 | 2,160 | +35% | 0 | 0 | — |
case-05 | pass→pass | 9,887 | 10,319 | +4% | 1 | 1 | 0% | 1,677 | 2,832 | +69% | 0 | 0 | — |
case-06 | pass→pass | 7,495 | 13,600 | +81% | 1 | 1 | 0% | 1,275 | 1,948 | +53% | 0 | 0 | — |
case-07 | pass→pass | 19,291 | 9,469 | -51% | 1 | 1 | 0% | 3,346 | 2,657 | -21% | 0 | 0 | — |
case-08 | fail→pass | 13,853 | 11,060 | -20% | 1 | 1 | 0% | 2,293 | 2,820 | +23% | 0 | 0 | — |
case-09 | fail→fail | 9,359 | 4,892 | -48% | 1 | 1 | 0% | 1,500 | 1,987 | +32% | 0 | 0 | — |
case-10 | pass→pass | 8,676 | 5,803 | -33% | 1 | 1 | 0% | 1,455 | 1,949 | +34% | 0 | 0 | — |
case-15 | pass→pass | 16,587 | 11,359 | -32% | 1 | 1 | 0% | 3,024 | 2,957 | -2% | 0 | 0 | — |
case-11 | fail→fail | 9,083 | 9,982 | +10% | 1 | 1 | 0% | 1,373 | 2,566 | +87% | 0 | 0 | — |
case-12 | fail→pass | 12,050 | 7,685 | -36% | 1 | 1 | 0% | 1,829 | 2,276 | +24% | 0 | 0 | — |
case-13 | fail→fail | 9,087 | 6,608 | -27% | 1 | 1 | 0% | 1,388 | 1,984 | +43% | 0 | 0 | — |
case-14 | pass→pass | 16,340 | 15,997 | -2% | 1 | 1 | 0% | 2,553 | 3,670 | +44% | 0 | 0 | — |
case-16 | pass→pass | 9,549 | 8,926 | -7% | 1 | 1 | 0% | 1,592 | 2,470 | +55% | 0 | 0 | — |
case-17 | pass→pass | 10,076 | 32,521 | +223% | 1 | 1 | 0% | 1,725 | 2,637 | +53% | 0 | 0 | — |
case-18 | fail→fail | 7,588 | 7,484 | -1% | 1 | 1 | 0% | 1,230 | 2,260 | +84% | 0 | 0 | — |
case-19 | pass→pass | 8,912 | 8,368 | -6% | 1 | 1 | 0% | 1,466 | 2,358 | +61% | 0 | 0 | — |
case-20 | pass→pass | 8,430 | 5,090 | -40% | 1 | 1 | 0% | 1,241 | 1,896 | +53% | 0 | 0 | — |
case-21 | fail→pass | 3,415 | 5,474 | +60% | 1 | 1 | 0% | 499 | 1,903 | +281% | 0 | 0 | — |
case-22 | pass→pass | 7,960 | 4,138 | -48% | 1 | 1 | 0% | 1,394 | 1,711 | +23% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +23 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.