Language tutor
Structured language tutor with adaptive pacing, placement testing, and granular grading for any language.
Free
Works with the AI Agents you already use
Language tutor
Example session with this skill installed
I'm ready for today's Spanish lesson. I want to continue where we left off with past tense verbs.
- Read your context and instructions
- Compiled the language tutor
Skill response
¡Hola! Let's keep going with the past tense.
Quick reminder: accents in the preterite
Last time trabaje slipped in instead of
trabajé. In the preterite the stress falls on the ending, so
yo -é and
él -ó always carry an accent. Without it, trabaje is a different form altogether.
- trabajé = I worked · trabajó = he/she worked
- Usual trap: pagar → pagué (the u keeps the hard g sound).
New: -er and -ir verbs, first two forms
Good news: -er and -ir verbs share the same past endings. Today we learn just two of them:
| comer | vivir | |
|---|---|---|
| yo | comí | viví |
| él / ella / usted | comió | vivió |
Same logic as -ar: the stress is on the ending, so the accent is always written.
- Ayer comí en casa. — Yesterday I ate at home.
- Mi abuela vivió en Sevilla. — My grandmother lived in Seville.
- Como pan. / Comí pan. — I eat bread. / I ate bread. One accent changes the tense.
Exercises
Answer in one message, by number. "no sé" is a fine answer.
Translate into Spanish
- Yesterday I worked until eight.
- Did you buy the bread?
- They arrived late.
- I paid by card.
- We spoke with Ana yesterday.
Translate into English
6. Ayer trabajé en casa.
Which one is in the past?
7. como · comí
Translate into Spanish
8. Yesterday I ate at home.
9. He lived in Madrid.
10. I lived in Lisbon.
Find the error
11. Ayer Marta comí paella.
Translate into Spanish
12. Yesterday Pablo ate late and I ate early.
¡Vamos!
Learner
1 Ayer trabaje hasta las ocho
2 ¿Compraste el pan?
3 Llegaron tarde
4 Pague con tarjeta
5 Hablamos con Ana ayer
6 Yesterday I worked at home
7 comí
8 Ayer comí en casa
9 Vivio en Madrid
10 no sé
11 comió
12 Ayer Pablo comió tarde y yo comí temprano
Skill response
| # | Your answer | Result | Correct form |
|---|---|---|---|
| 1 | Ayer trabaje hasta las ocho | ½ | Ayer trabajé hasta las ocho. |
| 2 | ¿Compraste el pan? | ✓ | |
| 3 | Llegaron tarde | ✓ | |
| 4 | Pague con tarjeta | ½ | Pagué con tarjeta. |
| 5 | Hablamos con Ana ayer | ✓ | |
| 6 | Yesterday I worked at home | ✓ | |
| 7 | comí | ✓ | |
| 8 | Ayer comí en casa | ✓ | |
| 9 | Vivio en Madrid | ✓ | Vivió en Madrid. |
| 10 | no sé | ✗ | Viví en Lisboa. |
| 11 | comió | ✓ | Ayer Marta comió paella. |
| 12 | Ayer Pablo comió tarde y yo comí temprano | ✓ |
10 out of 12. Good start on the new verbs.
1 and 4: the same mechanism as last time. trabaje and pague are real words with a different meaning, so the accent matters here. This is the one thing to watch: yo in the past =
-é.
- 10: viví is comí with the -ir verb. Same ending, same accent.
Remember today
- -er and -ir verbs share past endings: comí / comió, viví / vivió.
- yo -é / -í and él -ó / -ió always carry the accent.
- pagar → pagué.
Next time: the remaining persons of comer and vivir. ¡Hasta mañana!
Connects securely to your tools. The creator never sees your data.
About this skill
Language learning fails when lessons are too fast, grading is binary, or progress is lost between sessions. This skill implements a structured pedagogical framework for language instruction, focusing on small-step material delivery and adaptive pacing. It handles everything from initial placement tests to granular grading with partial credit, ensuring learners stay in the "Goldilocks zone" of difficulty.
What it does
- Structured Onboarding conducts self-assessments and placement tests to establish a baseline level.
- Adaptive Pacing selects lesson modes based on historical performance data and spaced repetition principles.
- Granular Grading awards partial credit for minor slips that don't change meaning, focusing on linguistic mastery over typing accuracy.
- Progress Persistence integrates with memory servers or local notes to track course position, item scores, and schedules.
- Error Correction identifies recurring failures and switches teaching angles rather than repeating identical drills.
How it works
- Load State reads the learner's profile, history, and goals from a memory server or local markdown file.
- Review executes a brief refresher of previously learned items to ensure retention before moving forward.
- Teach delivers new material in small, digestible pieces in the target language.
- Exercise generates typed drills to test the new material and records results with granular scores.
Frameworks & tools
This skill is designed to work with any chat-based LLM and is optimized for use with the Language Tutor MCP server or local Markdown-based memory backends.
Why this beats prompting it yourself
Standard prompts often oscillate between overly simple drills and overwhelming walls of text. This skill enforces a "small steps" rule and a fair grading system that accounts for human error, preventing the frustration of binary "correct/incorrect" feedback in complex grammar scenarios.
Use cases
- Daily Practice generating a 10-minute refresher and one new grammar point.
- Placement Testing determining where a user should start in a new language course.
- Targeted Review focusing specifically on items the learner has failed in previous sessions.
- Automated Grading checking long-form translations and providing detailed feedback on partial errors.
Known limitations
Requires a manual progress record if a memory server is not configured. Grading is performed based on internal linguistic knowledge rather than external dictionary APIs.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.
Relies on session history for pacing accuracy. Grade disputes require manual overrides. Best results require an MCP memory server or structured markdown notes for persistence.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 7 days ago
- Free with an Agensi account
- Software
- Language Tutor MCP
Git
Works best with language memory MCP server, but does not require it.
Permissions
No special permissions declared or detected