2026-09-22 03:55:56 +03:00
$MODULE Unidad 1 — Llegada al aeropuerto
2026-09-22 00:09:22 +00:00
FORMAT: 2
DIOCO_DOC_ID: lc_es_verano_sevilla_u01
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
DESCRIPTION: Знакомство, национальности, глагол ser, числа 0-20, вопросительные слова, tú/usted, глагол estar.
2026-09-22 00:09:22 +00:00
TARGET_LANG_G: es
HOME_LANG_G: ru
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
VOICE_DEFAULT: Sulafat | Hable español peninsular con claridad, a ritmo de principiante
VOICE_INTRO: Kore | Дружелюбный рассказчик, спокойный темп
VOICE: Lara | Leda | Estudiante extranjera, joven, curiosa y algo tímida
VOICE: Sofia | Aoede | Sevillana joven, alegre y habladora
VOICE: Kenji | Puck | Estudiante japonés, joven, amable y tranquilo
VOICE: Diego | Charon | Estudiante joven, bromista y relajado
VOICE: Carmen | Gacrux | Señora mayor, la casera, cálida y algo formal
VOICE: Valentina | Callirrhoe | Joven mexicana, cercana y expresiva
VOICE: Narrator | Kore | Narrador neutro
2026-09-22 04:22:43 +03:00
$SESSION Unidad 1
2026-09-22 00:09:22 +00:00
$GRAMMAR Objetivos y personajes
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s1-objetivos
2026-09-22 00:09:22 +00:00
**Цели юнита**
- Здороваться, представляться и прощаться.
- Называть имя, национальность и откуда вы.
- Использовать глагол **ser** (быть) и личные местоимения.
- Считать от 0 до 20 и диктовать по буквам.
- Задавать вопросы: вопросительные слова (qué, quién, cómo, dónde, de dónde…).
- Различать обращение на **ты (tú)** и на **Вы (usted)**, спрашивать «как дела».
**Персонажи:** Lara (только что приехала), Diego (испанец, из Севильи), Sofía (итальянка), Kenji (японец); в конце юнита — первое сообщение от Valentina, лучшей подруги Lara, которая тем же летом учится в Ciudad de México.
2026-09-22 04:22:43 +03:00
$SESSION Lección 1 · ¡Bienvenida a Sevilla!
2026-09-22 00:09:22 +00:00
$DIALOGUE En el aeropuerto
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s2-dialogo-aeropuerto
2026-09-22 00:09:22 +00:00
INTRO: Lara только что прилетела в Севилью на летний курс испанского. Этим летом она будет жить в общей квартире с В
INSTRUCTION: Слушайте и повторяйте каждую реплику.
REPEAT
VOCAB: hola
2026-09-22 04:22:43 +03:00
VOCAB: sí
2026-09-22 00:09:22 +00:00
Lara: Hola. ¿Eres Diego?
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
VOCAB: encantado
VOCAB: bienvenido
2026-09-22 00:09:22 +00:00
Diego: Sí, soy Diego. ¿Tú eres Lara?
Lara: Sí, soy Lara. Encantada.
Diego: Encantado. Bienvenida a Sevilla.
VOCAB: el viaje
Diego: ¿Qué tal el viaje?
2026-09-22 04:22:43 +03:00
VOCAB: cansado
2026-09-22 00:09:22 +00:00
Lara: Bien, gracias. Un poco cansada.
VOCAB: la maleta
Diego: Normal. ¿Es tu maleta?
Lara: Sí, es mi maleta. Es muy grande.
VOCAB: vamos
Diego: ¡Vamos! El piso está muy cerca.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
$DIALOGUE Haz el papel de Lara
ID: u01-s2-dialogo-aeropuerto-role-lara
REFERENCE: u01-s2-dialogo-aeropuerto
PLAY_ROLE: Lara
INTRO: Тот же разговор, но теперь Lara — это вы. Говорите её реплики в паузах.
INSTRUCTION: Вы — Lara. Произносите её реплики.
2026-09-22 00:09:22 +00:00
$GRAMMAR Nota cultural — Dos besos
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s2-nota-besos
**Nota cultural — Dos besos.** При знакомстве в Испании женщины и пары «мужчина–женщина» часто здороваются двумя поцелуями в щёки (сначала левая). Двое мужчин обычно жмут руки. В
2026-09-22 00:09:22 +00:00
$GRAMMAR el verbo SER y los pronombres
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s2-gramatica-ser
2026-09-22 00:09:22 +00:00
Самый главный глагол для знакомства — **ser** (быть). С
**Pronombres personales (личные местоимения)**
| Español | Traducción |
|---|---|
| yo | я |
| tú | ты |
| él / ella | он / она |
| usted | Вы (вежливо) |
| nosotros / nosotras | мы |
| vosotros / vosotras | вы (мн., Испания) |
| ellos / ellas | они |
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
> В
2026-09-22 00:09:22 +00:00
**El verbo SER — presente** (глагол «быть» — настоящее время)
| Pronombre | SER | Traducción |
|---|---|---|
| yo | **soy** | я есть |
| tú | **eres** | ты есть |
| él / ella / usted | **es** | он / она / Вы есть |
| nosotros/-as | **somos** | мы есть |
| vosotros/-as | **sois** | вы есть |
| ellos / ellas | **son** | они есть |
> Правильные, неправильные и отклоняющиеся глаголы подробно разобраны в 3 частях пособия «300 наиболее популярных глаголов испанского языка». Больше информации на сайте espanolenchanclas.com.
**¿Para qué lo usamos?** (для чего используем?) — имя, профессия, происхождение, описание «по сути»:
- {Soy Lara.} — Я Lara.
- {¿Tú eres Diego?} — Ты Diego?
- {Es mi maleta.} — Это мой чемодан.
- {Somos estudiantes.} — Мы студенты.
**Saludos y despedidas** (приветствия и прощания)
| Español | Traducción |
|---|---|
| Hola. | Привет. |
| Buenos días. | Доброе утро / добрый день. |
| Buenas tardes. | Добрый день (после обеда). |
| Buenas noches. | Добрый вечер / спокойной ночи. |
| Encantado. / Encantada. | Очень приятно. |
| Adiós. / Hasta luego. | Пока. / До скорого. |
$PRODUCE Ejercicio 1 · Practica el verbo SER
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s2-ej1-ser
INSTRUCTION: Поставьте глагол **ser** в нужную форму или переведите.
2026-09-22 00:09:22 +00:00
INPUT: type
CHECK: exact
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Yo ____{soy} Lara. (ser)
2026-09-22 00:09:22 +00:00
RESPONSE: Yo soy Lara.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Tú ____{eres} Diego. (ser)
2026-09-22 00:09:22 +00:00
RESPONSE: Tú eres Diego.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Él ____{es} estudiante. (ser)
2026-09-22 00:09:22 +00:00
RESPONSE: Él es estudiante.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Nosotros ____{somos} estudiantes. (ser)
2026-09-22 00:09:22 +00:00
RESPONSE: Nosotros somos estudiantes.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Ellos ____{son} amigos. (ser)
2026-09-22 00:09:22 +00:00
RESPONSE: Ellos son amigos.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Ella ____{es} Sofía. (ser)
2026-09-22 00:09:22 +00:00
RESPONSE: Ella es Sofía.
TEMPLATE: Спросите: «Ты Lara?»
RESPONSE: ¿Tú eres Lara?
TEMPLATE: Скажите: «Это мой чемодан.»
RESPONSE: Es mi maleta.
TEMPLATE: «Вы (вежливо) Diego?»
RESPONSE: ¿Usted es Diego?
TEMPLATE: «Мы студенты, а
RESPONSE: Nosotros somos estudiantes y ellos son amigos.
2026-09-22 04:22:43 +03:00
$SESSION Lección 2 · ¿De dónde eres?
2026-09-22 00:09:22 +00:00
$DIALOGUE En el piso compartido
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s3-dialogo-piso
2026-09-22 00:09:22 +00:00
INTRO: Lara приезжает в квартиру. Там её ждут соседи: Sofía из Италии и Kenji из Японии.
INSTRUCTION: Слушайте и повторяйте каждую реплику.
REPEAT
VOCAB: el piso
VOCAB: la compañera
2026-09-22 04:22:43 +03:00
VOCAB: el compañero
2026-09-22 00:09:22 +00:00
VOCAB: nuevo
Sofia: ¡Hola! Yo soy Sofía. ¿Tú eres la nueva compañera?
Lara: Sí, soy Lara. Encantada.
VOCAB: ¿De dónde eres?
VOCAB: de Roma
2026-09-22 04:22:43 +03:00
VOCAB: de Tokio
VOCAB: de Sevilla
2026-09-22 00:09:22 +00:00
Sofia: Encantada. Yo soy italiana, de Roma. ¿Y tú? ¿De dónde eres?
Lara: Es un secreto. Pero hablo un poco de español.
Kenji: Hola, yo soy Kenji. Soy japonés, de Tokio.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
VOCAB: el español
VOCAB: hablar
2026-09-22 00:09:22 +00:00
Diego: Y yo soy Diego, español, de Sevilla. Aquí todos hablamos español.
Sofia: Somos compañeros de piso. ¡Y ahora somos amigos!
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
$DIALOGUE Haz el papel de Sofia
ID: u01-s3-dialogo-piso-role-sofia
REFERENCE: u01-s3-dialogo-piso
PLAY_ROLE: Sofia
INTRO: Тот же разговор, но теперь Sofía — это вы. Говорите её реплики в паузах.
INSTRUCTION: Вы — Sofía. Произносите её реплики.
$GRAMMAR nacionalidades
ID: u01-s3-gramatica-nacionalidades
2026-09-22 00:09:22 +00:00
Национальности меняются по роду.
**Regla general** (общее правило) — -o → -a
| Masculino | Femenino | Traducción |
|---|---|---|
| italiano | italiana | итальянец / итальянка |
| mexicano | mexicana | мексиканец / мексиканка |
| chino | china | китаец / китаянка |
**Con consonante final (+ a)** (если на конце согласная — добавляем -a)
| Masculino | Femenino | Traducción |
|---|---|---|
| español | española | испанец / испанка |
| japonés | japonesa | японец / японка |
| alemán | alemana | немец / немка |
**Igual en los dos géneros** (одинаково в мужском и женском роде)
| Forma única | Traducción |
|---|---|
| marroquí | марокканец / марокканка |
| estudiante | студент / студентка |
**Preguntar el origen** (спросить, откуда человек)
- {¿De dónde eres?} — Откуда ты?
- {Soy de Roma.} — Я из Рима.
- {Soy italiana.} — Я итальянка.
- {¿Eres español?} — Ты испанец?
2026-09-22 04:22:43 +03:00
$GRAMMAR las palabras interrogativas
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s3-gramatica-interrogativas
2026-09-22 00:09:22 +00:00
В с
| Español | Traducción | Пример |
|---|---|---|
| ¿qué? | что? / какой? | {¿Qué es esto?} |
| ¿quién? · ¿quiénes? | кто? (ед. · мн.) | {¿Quién es?} · {¿Quiénes son?} |
| ¿cómo? | как? | {¿Cómo te llamas?} |
| ¿dónde? | где? | {¿Dónde está el piso?} |
| ¿de dónde? | откуда? | {¿De dónde eres?} |
2026-09-22 04:22:43 +03:00
| ¿adónde? | куда? | ¿Adónde vas? *(Unidad 8)* |
| ¿cuándo? | когда? | ¿Cuándo llegas? *(Unidad 7)* |
| ¿cuánto/-a/-os/-as? | сколько? | ¿Cuántos años tienes? *(Unidad 4)* |
| ¿cuál? · ¿cuáles? | какой (из)? который? | ¿Cuál prefieres? *(Unidad 13)* |
2026-09-22 00:09:22 +00:00
| ¿por qué? | почему? | {¿Por qué estudias español?} |
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
> - **¿por qué? → porque.** Вопрос «почему?» — *¿por qué?* (раздельно, с
> - **¿cuánto?** согласуется с
> - **¿qué? или ¿cuál?** — *¿Qué* + существительное (*¿Qué color?*), а
2026-09-22 00:09:22 +00:00
> - Т е
> Подробные таблицы вопросительных слов и других тем — в материалах на сайте espanolenchanclas.com.
$PRODUCE Ejercicio 2 · Masculino → femenino
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s3-ej2-femenino
2026-09-22 00:09:22 +00:00
INSTRUCTION: Преобразуйте в женский род: «Ella es …».
INPUT: type
CHECK: exact
TEMPLATE: italiano → ____
RESPONSE: Ella es italiana.
TEMPLATE: español → ____
RESPONSE: Ella es española.
TEMPLATE: japonés → ____
RESPONSE: Ella es japonesa.
TEMPLATE: mexicano → ____
RESPONSE: Ella es mexicana.
TEMPLATE: alemán → ____
RESPONSE: Ella es alemana.
TEMPLATE: chino → ____
RESPONSE: Ella es china.
TEMPLATE: marroquí → ____
RESPONSE: Ella es marroquí.
TEMPLATE: argentino → ____
RESPONSE: Ella es argentina.
TEMPLATE: portugués → ____
RESPONSE: Ella es portuguesa.
TEMPLATE: estadounidense → ____
RESPONSE: Ella es estadounidense.
$PRODUCE Ejercicio 3 · Traducir
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s3-ej3-traducir
2026-09-22 04:22:43 +03:00
INSTRUCTION: Переведите на испанский.
INPUT: type
2026-09-22 00:09:22 +00:00
CHECK: llm
TEMPLATE: Откуда ты?
RESPONSE: ¿De dónde eres?
TEMPLATE: Я из Рима.
RESPONSE: Soy de Roma.
TEMPLATE: Я итальянка.
RESPONSE: Soy italiana.
TEMPLATE: Он японец.
RESPONSE: Él es japonés.
TEMPLATE: Мы соседи по квартире.
RESPONSE: Somos compañeros de piso.
TEMPLATE: Diego испанец, из Севильи.
RESPONSE: Diego es español, de Sevilla.
TEMPLATE: А
RESPONSE: ¡Y ahora somos amigos!
TEMPLATE: Они немцы.
RESPONSE: Ellos son alemanes.
TEMPLATE: Откуда вы? (мн. ч., Испания)
RESPONSE: ¿De dónde sois?
TEMPLATE: Мы не итальянцы, мы испанцы.
RESPONSE: No somos italianos, somos españoles.
$PRODUCE Ejercicio 4 · Las palabras interrogativas
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s3-ej4-interrogativas
2026-09-22 00:09:22 +00:00
INSTRUCTION: Вставьте подходящее вопросительное слово.
INPUT: type
CHECK: exact
2026-09-22 04:22:43 +03:00
TEMPLATE: ¿____{Cómo} te llamas? — Me llamo Lara.
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Cómo te llamas? — Me llamo Lara.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ACCEPT: Cómo
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: ¿De ____{dónde} eres? — De Roma.
2026-09-22 00:09:22 +00:00
RESPONSE: ¿De dónde eres? — De Roma.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ACCEPT: dónde
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: ¿____{Quién} es? — Es Diego.
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Quién es? — Es Diego.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ACCEPT: Quién
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: ¿____{Qué} es esto? — Es mi maleta.
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Qué es esto? — Es mi maleta.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ACCEPT: Qué
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: ¿____{Dónde} está el piso? — Está cerca.
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Dónde está el piso? — Está cerca.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ACCEPT: Dónde
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: ¿____{Qué} tal el viaje? — Bien, gracias.
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Qué tal el viaje? — Bien, gracias.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ACCEPT: Qué
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: ¿____{Quiénes} son ellos? — Son Sofía y Kenji. (кто, мн.)
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Quiénes son ellos? — Son Sofía y Kenji.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ACCEPT: Quiénes
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: ¿____{Cuántos} años tienes? — Veinte. (сколько)
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Cuántos años tienes? — Veinte.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ACCEPT: Cuántos
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: ¿____{Cuál} es tu maleta? — La roja. (которая из)
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Cuál es tu maleta? — La roja.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ACCEPT: Cuál
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: ¿____{Por qué} no contestas? — Porque estoy cansada. (почему)
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Por qué no contestas? — Porque estoy cansada.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ACCEPT: Por qué
2026-09-22 00:09:22 +00:00
$PRODUCE Ejercicio 4.1 · Las palabras interrogativas (traducir)
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s3-ej4-1-interrogativas-traducir
2026-09-22 00:09:22 +00:00
INSTRUCTION: Переведите на испанский.
2026-09-22 04:22:43 +03:00
INPUT: type
2026-09-22 00:09:22 +00:00
CHECK: llm
TEMPLATE: «Ты испанец?»
RESPONSE: ¿Eres español?
TEMPLATE: «Откуда вы?» (мн. ч., Испания)
RESPONSE: ¿De dónde sois?
TEMPLATE: «Почему? — Потому что да.»
RESPONSE: ¿Por qué? — Porque sí.
TEMPLATE: «Как тебя зовут?»
RESPONSE: ¿Cómo te llamas?
TEMPLATE: «Кто это?»
RESPONSE: ¿Quién es?
TEMPLATE: «Где квартира?»
RESPONSE: ¿Dónde está el piso?
TEMPLATE: «Как дела?»
RESPONSE: ¿Qué tal?
TEMPLATE: «Кто они?» (мн. ч.)
RESPONSE: ¿Quiénes son?
TEMPLATE: «Сколько тебе лет?»
RESPONSE: ¿Cuántos años tienes?
TEMPLATE: «Какой твой чемодан? — Красный.»
RESPONSE: ¿Cuál es tu maleta? — La roja.
2026-09-22 04:22:43 +03:00
$SESSION Lección 3 · Los números
2026-09-22 00:09:22 +00:00
$GRAMMAR los números (0-20)
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s4-gramatica-numeros
2026-09-22 00:09:22 +00:00
**Números 0-10**
| 0 | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|
| cero | uno | dos | tres | cuatro | cinco |
| 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|
| seis | siete | ocho | nueve | diez |
**Números 11-20**
| 11 | 12 | 13 | 14 | 15 |
|---|---|---|---|---|
| once | doce | trece | catorce | quince |
| 16 | 17 | 18 | 19 | 20 |
|---|---|---|---|---|
| dieciséis | diecisiete | dieciocho | diecinueve | veinte |
**Frases útiles (deletrear)**
Сам алфавит и названия букв — в **Unidad 0**. Здесь — полезные фразы, чтобы переспросить и продиктовать:
- {¿Cómo te llamas?} — Как тебя зовут?
- {¿Cómo se escribe?} — Как это пишется?
- {¿Puedes repetir, por favor?} — Можешь повторить, пожалуйста?
$PRODUCE Ejercicio 5 · Cuenta y deletrea
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s4-ej5-numeros
2026-09-22 00:09:22 +00:00
INSTRUCTION: Напишите словами по-испански.
2026-09-22 04:22:43 +03:00
INPUT: type
CHECK: exact
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: 7 → ____
2026-09-22 00:09:22 +00:00
RESPONSE: siete
2026-09-22 04:22:43 +03:00
TEMPLATE: 12 → ____
2026-09-22 00:09:22 +00:00
RESPONSE: doce
2026-09-22 04:22:43 +03:00
TEMPLATE: 20 → ____
2026-09-22 00:09:22 +00:00
RESPONSE: veinte
2026-09-22 04:22:43 +03:00
TEMPLATE: 16 → ____
2026-09-22 00:09:22 +00:00
RESPONSE: dieciséis
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Спросите: «Как это пишется?»
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Cómo se escribe?
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Попросите повторить, пожалуйста.
2026-09-22 00:09:22 +00:00
RESPONSE: ¿Puedes repetir, por favor?
2026-09-22 04:22:43 +03:00
TEMPLATE: 11 → ____
2026-09-22 00:09:22 +00:00
RESPONSE: once
2026-09-22 04:22:43 +03:00
TEMPLATE: 14 → ____
2026-09-22 00:09:22 +00:00
RESPONSE: catorce
2026-09-22 04:22:43 +03:00
TEMPLATE: 19 → ____
2026-09-22 00:09:22 +00:00
RESPONSE: diecinueve
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: «Меня зовут Ана. Как это пишется?»
2026-09-22 00:09:22 +00:00
RESPONSE: Me llamo Ana. ¿Cómo se escribe?
2026-09-22 04:22:43 +03:00
$SESSION Lección 4 · Tú y usted · ¿Cómo estás?
2026-09-22 00:09:22 +00:00
$DIALOGUE En la puerta
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s5-dialogo-puerta
2026-09-22 00:09:22 +00:00
INTRO: Lara saluda primero a doña Carmen, la casera (formal), y después a Diego (con confianza).
INSTRUCTION: Слушайте и повторяйте каждую реплику.
REPEAT
VOCAB: usted
2026-09-22 04:22:43 +03:00
VOCAB: el señor
2026-09-22 00:09:22 +00:00
VOCAB: la señora
2026-09-22 04:22:43 +03:00
VOCAB: el inquilino
2026-09-22 00:09:22 +00:00
VOCAB: la inquilina
Carmen: Buenos días. ¿Usted es la nueva inquilina?
Lara: Sí, soy Lara. Encantada, señora.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
VOCAB: el casero
VOCAB: la casera
2026-09-22 00:09:22 +00:00
Carmen: Mucho gusto. ¿Cómo está usted?
VOCAB: ¿Y usted?
2026-09-22 04:22:43 +03:00
VOCAB: ¿Y tú?
2026-09-22 00:09:22 +00:00
Lara: Muy bien, gracias. ¿Y usted?
Carmen: Bien, bien. Bienvenida al edificio.
2026-09-22 04:22:43 +03:00
Narrator: (Llega Diego.)
2026-09-22 00:09:22 +00:00
Diego: ¡Hola, Lara! ¿Qué tal? ¿Cómo estás?
VOCAB: genial
Lara: ¡Genial! Un poco cansada del viaje.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
$DIALOGUE Haz el papel de Lara
ID: u01-s5-dialogo-puerta-role-lara
REFERENCE: u01-s5-dialogo-puerta
PLAY_ROLE: Lara
INTRO: Тот же разговор, но теперь Lara — это вы. Говорите её реплики в паузах.
INSTRUCTION: Вы — Lara. Произносите её реплики.
$GRAMMAR tú y usted
ID: u01-s5-gramatica-tu-usted
2026-09-22 00:09:22 +00:00
В
- **tú** — неформально: друзья, ровесники, молодёжь, дети. В
- **usted** — вежливо/формально: незнакомые, пожилые люди, начальство, официальные ситуации.
Важно: **usted** идёт с е
| | tú (неформально) | usted (вежливо) |
|---|---|---|
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
| как зовут | ¿Cómo **te** llamas? | ¿Cómo **se** llama? |
| откуда | ¿De dónde **eres**? | ¿De dónde **es**? |
| как дела | ¿Cómo **estás**? | ¿Cómo **está**? |
2026-09-22 00:09:22 +00:00
> **Nota cultural.** В с б а р е В
$GRAMMAR el verbo ESTAR y «¿cómo estás?»
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s5-gramatica-estar
2026-09-22 00:09:22 +00:00
Чтобы спросить и сказать, **как дела / как самочувствие**, нужен глагол **estar**:
| Pronombre | estar |
|---|---|
| yo | estoy |
| tú | estás |
| él / ella / usted | está |
| nosotros/as | estamos |
| vosotros/as | estáis |
| ellos/as / ustedes | están |
> Правильные, неправильные и отклоняющиеся глаголы подробно разобраны в 3 частях пособия «300 наиболее популярных глаголов испанского языка». Больше информации на сайте espanolenchanclas.com.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
- {¿Cómo estás?} — Estoy **bien / mal / regular**.
2026-09-22 00:09:22 +00:00
- Estoy **cansada** (устала), **contento** (доволен), **nervioso** (нервничаю).
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
> **ser или estar?** Коротко: **ser** — какой человек вообще (*soy alta* — я высокая), **estar** — как он себя чувствует сейчас (*estoy cansada* — я устала). Про **estar** в значении «где находится» поговорим в следующем юните.
2026-09-22 00:09:22 +00:00
$PRODUCE Ejercicio 6 · Tú o usted (traducir)
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s5-ej6-tu-usted
INSTRUCTION: Переведите на испанский, выбирая правильное обращение.
2026-09-22 04:22:43 +03:00
INPUT: type
2026-09-22 00:09:22 +00:00
CHECK: llm
TEMPLATE: (вежливо) Как В а с
RESPONSE: ¿Cómo se llama (usted)?
TEMPLATE: (неформально) Откуда ты?
RESPONSE: ¿De dónde eres?
TEMPLATE: Как ты? — Я устал.
RESPONSE: ¿Cómo estás? — Estoy cansado.
TEMPLATE: (вежливо) Как Вы?
RESPONSE: ¿Cómo está usted?
TEMPLATE: Я отлично, но немного устала.
RESPONSE: Estoy genial, pero un poco cansada.
TEMPLATE: (неформально) Как тебя зовут?
RESPONSE: ¿Cómo te llamas?
TEMPLATE: (вежливо) Откуда Вы?
RESPONSE: ¿De dónde es usted?
TEMPLATE: Я нервничаю.
RESPONSE: Estoy nervioso/a.
TEMPLATE: (вежливо) Как Вы? Вы устали?
RESPONSE: ¿Cómo está usted? ¿Está cansado/a?
TEMPLATE: Мы в порядке, спасибо.
RESPONSE: Estamos bien, gracias.
2026-09-22 04:22:43 +03:00
$SESSION Lectura y comprensión
2026-09-22 00:09:22 +00:00
$DIALOGUE En el aeropuerto
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s6-lectura
2026-09-22 00:09:22 +00:00
INTRO: Lara только что прилетела. Прочитай, как проходит её первая встреча в аэропорту Севильи.
2026-09-22 04:22:43 +03:00
Narrator: Lara llega al aeropuerto de Sevilla.
LINE: Es por la tarde y hace calor.
LINE: Lara está un poco nerviosa, pero está muy contenta.
LINE: En la puerta hay un chico con un cartel: «LARA».
LINE: — ¡Hola! ¿Eres tú Lara? —pregunta el chico.
LINE: — Sí, soy yo —responde Lara.
LINE: — Yo soy Diego. Soy español, de Sevilla. ¡Encantado!
LINE: Con Diego están dos amigos: Sofía y Kenji.
LINE: Sofía es italiana; es de Roma.
LINE: Kenji es japonés; es de Tokio.
LINE: Los tres son estudiantes y son muy simpáticos.
LINE: — ¿De dónde eres, Lara? —pregunta Sofía.
LINE: — Es un secreto —dice Lara con una sonrisa.
LINE: Ahora son las cuatro de la tarde.
LINE: Lara no está sola: tiene tres amigos nuevos.
LINE: Su verano en Sevilla empieza hoy.
2026-09-22 00:09:22 +00:00
$SELECT Comprensión · Elige la opción correcta
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s6-comprension-a
REFERENCE: u01-s6-lectura
2026-09-22 04:22:43 +03:00
INSTRUCTION: Выберите правильный вариант.
2026-09-22 00:09:22 +00:00
SHOW_PROMPT
PROMPT: Lara llega a Sevilla por la…
OPTION: a | mañana
OPTION: b | tarde
OPTION: c | noche
ANSWER: b
PROMPT: Diego es…
OPTION: a | italiano
OPTION: b | japonés
OPTION: c | español
ANSWER: c
PROMPT: Sofía es de…
OPTION: a | Roma
OPTION: b | Tokio
OPTION: c | Sevilla
ANSWER: a
PROMPT: Kenji es…
OPTION: a | español
OPTION: b | japonés
OPTION: c | italiano
ANSWER: b
PROMPT: Sobre su país, Lara dice…
OPTION: a | que es de Italia
OPTION: b | que es un secreto
OPTION: c | que es de Japón
ANSWER: b
$PRODUCE Comprensión · Responde a las preguntas
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s6-comprension-b
REFERENCE: u01-s6-lectura
2026-09-22 04:22:43 +03:00
INSTRUCTION: Ответьте на вопросы по тексту.
INPUT: speak
2026-09-22 00:09:22 +00:00
CHECK: llm
SHOW_PROMPT
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
RUBRIC: Ключ — факт из текста; принять любую формулировку с
2026-09-22 00:09:22 +00:00
PROMPT: ¿Cómo se llama el chico del aeropuerto?
RESPONSE: Se llama Diego.
PROMPT: ¿De dónde es Sofía?
RESPONSE: Es de Roma (italiana).
PROMPT: ¿Qué hora es?
RESPONSE: Son las cuatro de la tarde.
PROMPT: ¿Cuántos amigos nuevos tiene Lara?
RESPONSE: Tres.
$DIALOGUE Mensaje de Valentina
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s6-mensaje-valentina
INTRO: Вечером Lara получает сообщение. Valentina — её лучшая подруга из той же страны: этим летом она тоже учит испанский, но выбрала курс в Ciudad de México. Девушки будут переписываться всё лето.
2026-09-22 00:09:22 +00:00
Valentina: ¡Hola, Lara! Soy Valentina, ¡tu amiga de siempre! ¿Qué tal Sevilla? Yo estoy en Ciudad de México. Mi departamento es grande y mis compañeros son muy simpáticos: Facundo es argentino y Camila es colombiana. Estoy un poco cansada, pero muy contenta. ¿Y tú, cómo estás? Un abrazo.
Lara: ¡Hola, Vale! Estoy muy bien. Sevilla es genial. Mis compañeros son Diego, Sofía y Kenji. ¡Un abrazo grande!
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
NOTES: **Ojo — departamento.** Valentina пишет *departamento*, а
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
$SESSION Lección 5 · Repaso y tarea final
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
$PRODUCE Ejercicio 7 · Repaso de la Unidad 1 (traducir)
ID: u01-s7-ej7-repaso
2026-09-22 04:22:43 +03:00
INSTRUCTION: Переведите на испанский.
INPUT: type
2026-09-22 00:09:22 +00:00
CHECK: llm
TEMPLATE: Привет, я Lara.
RESPONSE: Hola, soy Lara.
TEMPLATE: Очень приятно (жен.).
RESPONSE: Encantada.
TEMPLATE: Откуда ты?
RESPONSE: ¿De dónde eres?
TEMPLATE: Я японец, из Токио.
RESPONSE: Soy japonés, de Tokio.
TEMPLATE: Это мой чемодан. Он очень большой.
RESPONSE: Es mi maleta. Es muy grande.
TEMPLATE: Мы соседи по квартире, а
RESPONSE: Somos compañeros de piso, y ahora amigos.
TEMPLATE: Добро пожаловать в Севилью!
RESPONSE: ¡Bienvenida a Sevilla!
TEMPLATE: Доброе утро! Как дела?
RESPONSE: ¡Buenos días! ¿Qué tal?
TEMPLATE: Извините, Вы новая жилица?
RESPONSE: Perdone, ¿es usted la nueva inquilina?
TEMPLATE: Я студентка испанского, и этим летом я в Севилье.
RESPONSE: Soy estudiante de español y este verano estoy en Sevilla.
$PRODUCE Tarea final · Tu presentación
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s7-tarea-final
INTRO: Запишите голосом (или напишите) презентацию о с е б е
INSTRUCTION: Представьтесь, назовите национальность, скажите, что вы студент(ка) испанского.
2026-09-22 04:22:43 +03:00
INPUT: type
2026-09-22 00:09:22 +00:00
CHECK: llm
2026-09-22 04:22:43 +03:00
LONGFORM
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Включите: 1) приветствие, 2) имя, 3) национальность / откуда вы, 4) что вы студент(ка) испанского. · Hola, me llamo ____. · Soy ____ . / Soy de ____ . · Soy estudiante de español. Estoy en Sevilla este verano.
2026-09-22 00:09:22 +00:00
RESPONSE: Hola, me llamo Lara. Soy estudiante de español. Encantada. Estoy en Sevilla este verano.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
RUBRIC: Принять любой связный текст с
2026-09-22 00:09:22 +00:00
$PRODUCE Mini-test
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s7-minitest
INSTRUCTION: Ответьте на вопросы мини-теста.
2026-09-22 04:22:43 +03:00
INPUT: type
2026-09-22 00:09:22 +00:00
CHECK: llm
TEMPLATE: Forma de ser para «yo».
RESPONSE: soy
TEMPLATE: Forma de ser para «nosotros».
RESPONSE: somos
TEMPLATE: Femenino de «español».
RESPONSE: española
TEMPLATE: El número 17.
RESPONSE: diecisiete
TEMPLATE: ¿Cómo se pregunta «Откуда ты?»
RESPONSE: ¿De dónde eres?
TEMPLATE: «Очень приятно» (жен.).
RESPONSE: Encantada.
TEMPLATE: «Как Вы?» (вежливо).
RESPONSE: ¿Cómo está usted?
2026-09-22 04:22:43 +03:00
$SESSION Vocabulario adicional
2026-09-22 00:09:22 +00:00
$DIALOGUE Nacionalidades y países
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s8-vocab-nacionalidades
2026-09-22 04:22:43 +03:00
INTRO: Необязательные слова по теме — учите по желанию.
2026-09-22 00:09:22 +00:00
REPEAT
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
LINE: francés — Francia
LINE: francesa — Francia
LINE: inglés — Inglaterra
LINE: inglesa — Inglaterra
LINE: alemán — Alemania
LINE: alemana — Alemania
LINE: portugués — Portugal
LINE: portuguesa — Portugal
LINE: ruso — Rusia
LINE: rusa — Rusia
LINE: estadounidense — Estados Unidos
LINE: brasileño — Brasil
LINE: brasileña — Brasil
LINE: argentino — Argentina
LINE: argentina — Argentina
LINE: mexicano — México
LINE: mexicana — México
LINE: chino — China
LINE: china — China
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
$DIALOGUE En el aeropuerto (vocabulario)
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s8-vocab-aeropuerto
2026-09-22 04:22:43 +03:00
INTRO: Необязательные слова по теме — учите по желанию.
2026-09-22 00:09:22 +00:00
REPEAT
LINE: la maleta
LINE: el equipaje
LINE: el pasaporte
LINE: el billete
LINE: la tarjeta de embarque
LINE: el vuelo
LINE: el avión
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
LINE: la puerta (de embarque)
2026-09-22 00:09:22 +00:00
LINE: la llegada
LINE: la salida
$DIALOGUE Saludos y cortesía
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s8-vocab-saludos
2026-09-22 04:22:43 +03:00
INTRO: Необязательные слова по теме — учите по желанию.
2026-09-22 00:09:22 +00:00
REPEAT
LINE: ¿Qué tal?
LINE: ¿Cómo estás?
LINE: Mucho gusto.
LINE: Igualmente.
LINE: Por favor.
LINE: Gracias.
LINE: De nada.
LINE: Perdón.
LINE: Perdona.
LINE: Perdone.
LINE: Hasta mañana.
LINE: Hasta pronto.
$DIALOGUE Estados y sensaciones
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
ID: u01-s8-vocab-estados
2026-09-22 04:22:43 +03:00
INTRO: Необязательные слова по теме — учите по желанию.
2026-09-22 00:09:22 +00:00
REPEAT
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
LINE: bien
LINE: mal
LINE: regular
2026-09-22 00:09:22 +00:00
LINE: cansado
LINE: contento
LINE: nervioso
LINE: enfermo
LINE: ocupado
LINE: aburrido
LINE: preocupado
LINE: tener sueño
LINE: tener hambre
LINE: tener sed
