alenakrauchanka/verano-en-sevilla
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
main
verano-en-sevilla/05.module

950 lines
45 KiB
Plaintext
Raw Permalink Normal View History

Move module titles onto the \ line (production parser no longer reads a TITLE: field)
2026-09-22 03:55:56 +03:00
$MODULE Unidad 5 — Paseo léxico · Quién soy: datos, etapas de la vida y carácter
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
FORMAT: 2
DIOCO_DOC_ID: lc_es_verano_sevilla_u05
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: Анкеты и личные данные, семья и этапы жизни, характер, праздники.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
TARGET_LANG_G: es
HOME_LANG_G: ru
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: Valentina | Callirrhoe | Joven mexicana, cercana y expresiva
VOICE: Narrator | Kore | Narrador neutro
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: Secretaria | Umbriel | Secretaria
VOICE: Elena | Algieba | Profesora
VOICE: Marina | Erinome | Amiga de Diego
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
$SESSION Unidad 5
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$GRAMMAR Objetivos y personajes
**Цели юнита**
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
- Понимать поля простой анкеты (*formulario*) и заполнять их: {nombre, apellidos, fecha y lugar de nacimiento, dirección, código postal, número de teléfono, firma}.
- Отвечать на вопросы о личных данных в секретариате, банке или отеле и получать документ: {tarjeta de estudiante}.
- Рассказывать о семейном положении и этапах жизни: {es soltera, está casado, tiene gemelos, es viudo}; {niño} → {adolescente} → {joven} → {adulto} → {mayor}.
- Описывать характер человека парами антонимов: {simpático / antipático, abierto / reservado, tranquilo / nervioso}.
- Поздравлять и говорить о празднике: {¡Felicidades!}, {¡Enhorabuena!}, {fiesta de cumpleaños, regalo, tarta, velas}.
- Называть людей вокруг: {compañero de clase / de piso / de trabajo, vecino, socio, conocido}.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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
**Грамматика:** только повторение — {ser} для имени, происхождения, профессии и черт характера *(Unidad 1, 4)*; {estar} + прилагательное для состояния *(Unidad 1)*; {tener} + возраст и {tener} + члены семьи *(Unidad 4)*; {hay} *(Unidad 2)*; согласование прилагательных и посессивы *(Unidad 4)*; числа 0–100, месяцы и дата *(Unidad 4)*; {tú / usted}, {me llamo / se llama}, вопросительные слова *(Unidad 1)*. Глаголы {nacer, casarse, invitar, regalar, felicitar} пока только как слова в словаре и в фиксированных выражениях — спряжение начинается с *Unidad 7*.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
**Персонажи:** Lara, Diego, Sofía, Kenji, Marina, la profesora Elena, la secretaria de la academia.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
$SESSION Lección 1 · El formulario
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$DIALOGUE En la secretaría
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
ID: u05-s1-dialogo-secretaria
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INTRO: Первый день занятий. Перед уроком Lara идёт в секретариат (la secretaría) школы: нужно заполнить анкету и получить студенческий билет. Secretaria говорит с ней на usted. Valentina в Ciudad de México на этой же неделе заполняет такой же formulario в своей школе.
INSTRUCTION: Слушайте и повторяйте каждую реплику.
REPEAT
VOCAB: el formulario
VOCAB: el impreso
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: rellenar
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: el nombre
VOCAB: los apellidos
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: apellidarse
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
Secretaria: Buenos días. ¿Es usted la estudiante nueva del curso de verano? Este es el formulario para la tarjeta de estudiante. Primero: ¿cómo se llama y cómo se apellida?
Lara: Buenos días. Me llamo Lara. Mi apellido es un poco largo… Aquí está, en el pasaporte.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: la fecha de nacimiento
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 lugar de nacimiento
VOCAB: nacer
VOCAB: el nacimiento
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
Secretaria: Perfecto, gracias. ¿Cuál es su fecha de nacimiento?
Lara: El 12 de abril de 2006. Tengo veinte años.
VOCAB: la nacionalidad
Secretaria: ¿Lugar de nacimiento y nacionalidad?
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: el sexo
VOCAB: el estado civil
VOCAB: el carné de identidad
VOCAB: el pasaporte
Lara: Eso es un secreto… pero también está en el pasaporte. Sexo: mujer. Estado civil: soltera. ¿Es necesario el carné de identidad?
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
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
VOCAB: es necesario
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: es suficiente
VOCAB: la dirección
VOCAB: el código postal
VOCAB: el número de teléfono
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Secretaria: No, con el pasaporte es suficiente. ¿Dirección en Sevilla, código postal y número de teléfono?
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Lara: Calle Betis, 14, segundo piso; código postal 41010. Y mi número es el 612 34 56 78.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: la firma
VOCAB: firmar
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: la tarjeta de estudiante
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Secretaria: Gracias. Su firma aquí, por favor… ¡Perfecto! Esta es su tarjeta de estudiante. ¡Bienvenida a la academia!
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Lara: ¡Muchas 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
$DIALOGUE Haz el papel de Lara
ID: u05-s1-dialogo-secretaria-role-lara
REFERENCE: u05-s1-dialogo-secretaria
PLAY_ROLE: Lara
INTRO: Тот же разговор, но теперь Lara — это вы. Говорите её реплики в паузах.
INSTRUCTION: Вы — Lara. Произносите её реплики.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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
$GRAMMAR Léxico en contexto: cómo leer un formulario
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Анкеты в школе, банке, отеле или на почте устроены одинаково: слева — название поля, справа — место для ответа. Поля почти всегда стоят в одном и том же порядке, поэтому достаточно один раз понять, что куда писать.
| Campo (поле) | Что писать | Ejemplo |
|---|---|---|
| Nombre | только имя | {Lara} |
| Apellidos | фамилия (у испанцев — две, поэтому поле во мн. ч.) | {García López} |
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
| Fecha de nacimiento | дата цифрами: день / месяц / год | {12/04/2006} — вслух {el 12 de abril de 2006} |
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
| Lugar de nacimiento | город (страна) | {Roma (Italia)} |
| Nacionalidad | прилагательное национальности *(Unidad 1)* | {italiana} |
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
| Sexo | {hombre (H) / mujer (M)} | {M} |
| Estado civil | {soltero/a, casado/a…} *(Lección 2)* | {soltera} |
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
| Dirección | улица, номер дома, этаж | {calle Betis, 14, 2.º} |
| Código postal | пять цифр | {41010} |
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
| Número de teléfono | девять цифр, читают парами | {612 34 56 78} — {seis-doce, treinta y cuatro, cincuenta y seis, setenta y ocho} |
| DNI / Pasaporte | номер документа | {X1234567} |
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
| Firma | подпись от руки | ✍ |
Слова анкеты образуют семьи: глагол → существительное → поле в бланке.
| Verbo | Sustantivo | En el formulario |
|---|---|---|
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
| nacer (родиться) | el nacimiento | {fecha / lugar de nacimiento} |
| firmar (подписывать) | la firma | {Firma: ______} |
| llamarse (называться) | el nombre | {Nombre: Lara} |
| apellidarse (носить фамилию) | el apellido | {Apellidos: García López} |
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
| rellenar (заполнять) | el formulario / el impreso | {rellenar el formulario} |
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
> **Ojo — nombre, apellido, nombre completo.** *Nombre* — это только имя. Фамилия — *apellido*, и у испанцев их две *(Unidad 4)*, поэтому в анкетах поле называется *Apellidos*. Имя + фамилии вместе — {el nombre completo}. Вопрос о фамилии строится как вопрос об имени: {¿Cómo se llama?} → {¿Cómo se apellida?} — {Me apellido García.} Если фамилию не понимают, выручает {¿Cómo se escribe?} *(Unidad 1)*.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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
> **Ojo — la fecha de nacimiento.** В бланке дату пишут цифрами ({12/04/2006}), вслух читают как обычную дату: {el 12 de abril de 2006} *(Unidad 4)*. Год пишем цифрами — большие числа появятся в следующих юнитах. Сказать «я родился(-ась) в…» пока можно фиксированной фразой: {Nací en 2006} / {Nací en Roma}; о другом человеке — {Sofía nació en Roma} *(Unidad 27)*.
> **Nota cultural — El DNI.** У каждого гражданина Испании есть **DNI** (*Documento Nacional de Identidad*) — пластиковая карточка с фотографией, номером и буквой. В разговоре говорят {el DNI} или {el carné}. Его просят везде: в банке, на почте, в отеле, при покупке билета на поезд. Иностранцы вместо DNI показывают {el pasaporte}, а те, кто живёт в Испании долго, получают **NIE** — *Número de Identidad de Extranjero*.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$PRODUCE Ejercicio 1 · Los campos del formulario
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
ID: u05-s1-ej1-campos
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INSTRUCTION: Дополните название поля анкеты (в скобках — подсказка).
INPUT: type
CHECK: exact
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: ____{Nombre}: Lara (имя)
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
RESPONSE: Nombre
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: ____{Apellidos}: García López (фамилии)
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
RESPONSE: Apellidos
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: ____{Fecha} de nacimiento: 12/04/2006 (дата)
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
RESPONSE: Fecha
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: ____{Lugar} de nacimiento: Roma (место)
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
RESPONSE: Lugar
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: ____{Nacionalidad}: italiana (гражданство)
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
RESPONSE: Nacionalidad
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Estado ____{civil}: soltera (семейное положение)
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
RESPONSE: civil
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: ____{Código} postal: 41010 (почтовый индекс)
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
RESPONSE: Código
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Número de ____{teléfono}: 612 34 56 78 (телефон)
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
RESPONSE: teléfono
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: ____{Carné} de identidad (DNI) o ____{pasaporte}: X1234567 (удостоверение; паспорт)
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
RESPONSE: Carné de identidad (DNI) o pasaporte
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: ____{Firma}: Lara — y aquí, su ____{tarjeta} de estudiante. (подпись; карточка, билет)
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
RESPONSE: Firma: Lara — y aquí, su tarjeta de estudiante.
$PRODUCE Ejercicio 1.1 · Los campos del formulario (traducir)
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
ID: u05-s1-ej1-1-traducir
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INSTRUCTION: Переведите на испанский.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
INPUT: type
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
CHECK: llm
TEMPLATE: Как Ваша фамилия? (usted)
RESPONSE: ¿Cómo se apellida usted?
TEMPLATE: Меня зовут Lara, а фамилия у меня длинная.
RESPONSE: Me llamo Lara y mi apellido es largo.
TEMPLATE: Какая у Вас дата рождения?
RESPONSE: ¿Cuál es su fecha de nacimiento?
TEMPLATE: Двенадцатое апреля 2006 года.
RESPONSE: El 12 de abril de 2006.
TEMPLATE: Место рождения: Рим. Гражданство: итальянка.
RESPONSE: Lugar de nacimiento: Roma. Nacionalidad: italiana.
TEMPLATE: Какой у тебя номер телефона?
RESPONSE: ¿Cuál es tu número de teléfono?
TEMPLATE: Мой адрес — улица Betis, 14. Почтовый индекс — 41010.
RESPONSE: Mi dirección es calle Betis, 14. El código postal es 41010.
TEMPLATE: Ваша подпись здесь, пожалуйста.
RESPONSE: Su firma aquí, por favor.
TEMPLATE: Это анкета для студенческого билета: имя, фамилии, дата и место рождения.
RESPONSE: Este es el formulario para la tarjeta de estudiante: nombre, apellidos, fecha y lugar de nacimiento.
TEMPLATE: Удостоверение личности не обязательно: паспорта достаточно.
RESPONSE: El carné de identidad no es necesario: con el pasaporte es suficiente.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
$SESSION Lección 2 · Estado civil y etapas de la vida
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$DIALOGUE La familia de Diego
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
ID: u05-s2-dialogo-familia
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INTRO: После занятий ребята пьют кофе в патио школы. Lara показывает новый студенческий билет, вспоминает графу «estado civil» — и разговор переходит на семью Diego.
INSTRUCTION: Слушайте и повторяйте каждую реплику.
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
VOCAB: soltero
VOCAB: soltera
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
Lara: Esta es mi tarjeta de estudiante. Y en el formulario hay una pregunta: «estado civil». Yo soy soltera, claro.
VOCAB: la pareja
Sofia: Yo también soy soltera, pero tengo pareja: mi novio se llama Luca y está en Roma.
VOCAB: divorciado
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: divorciada
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: divorciarse
Diego: Pues mi familia es un poco complicada. Mis padres están divorciados. Mi madre tiene pareja nueva y mi padre está soltero otra vez.
Kenji: ¿Y tus hermanos?
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: casado
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: casada
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: casarse
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: los gemelos
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: las gemelas
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: el niño
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: la niña
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
Diego: Mi hermano mayor está casado. Su mujer se llama Rocío y tienen gemelos: dos niños de dos años.
Lara: ¡Gemelos! ¿Y tus abuelos?
VOCAB: viudo
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: viuda
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: el adolescente
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: la adolescente
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
Diego: Mi abuelo es viudo. Tiene ochenta y dos años, pero es muy activo. Y mi hermana pequeña es adolescente: tiene catorce años.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: el joven
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: la joven
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: el adulto
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: la adulta
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: mayor
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: anciano
VOCAB: anciana
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Sofia: Entonces en tu familia hay niños, una adolescente, jóvenes, adultos y un abuelo mayor. ¡Todas las etapas de la vida!
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: el ex
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: la ex
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Diego: Sí. Y también hay un ex: el ex marido de mi tía. Pero eso es otra historia.
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 Diego
ID: u05-s2-dialogo-familia-role-diego
REFERENCE: u05-s2-dialogo-familia
PLAY_ROLE: Diego
INTRO: Тот же разговор, но теперь Diego — это вы. Говорите его реплики в паузах.
INSTRUCTION: Вы — Diego. Произносите его реплики.
$GRAMMAR Léxico en contexto: la línea de la vida y el estado civil
Испанский делит жизнь на понятные этапы (*las etapas de la vida*), и у каждого этапа есть своё слово. Возраст при этом всегда через {tener}: {tiene dos años}, {tiene ochenta y dos años} *(Unidad 4)*.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
| Etapa | Edad (примерно) | Ejemplo |
|---|---|---|
| el niño / la niña | 2–12 | {Los gemelos son niños pequeños.} |
| el/la adolescente | 13–17 | {La hermana de Diego es adolescente.} |
| el/la joven | 18–30 | {Lara y Sofía son jóvenes.} |
| el adulto / la adulta | 30–65 | {Los padres de Diego son adultos.} |
| mayor · anciano/a | 65+ | {El abuelo es mayor.} |
Семейное положение (*el estado civil*) — четыре слова из анкеты. С ними по-разному ведут себя *ser* и *estar*:
| Estado civil | Verbo | Ejemplo |
|---|---|---|
| soltero/a | **ser** (или *estar* — «сейчас без пары») | {Lara es soltera.} |
| casado/a | **estar** | {Mi hermano está casado.} |
| divorciado/a · separado/a | **estar** | {Mis padres están divorciados.} |
| viudo/a | **ser** | {Mi abuelo es viudo.} |
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
Логика простая: {casado, divorciado, separado} — результат события (свадьбы, развода), поэтому {estar}; {soltero} и {viudo} — «категория», поэтому {ser}. Семьи слов: {casarse con alguien} → {casado/a} → {la boda} (свадьба); {divorciarse} → {divorciado/a}; {nacer} → {el nacimiento}; {crecer} — расти: из {niño} в {adulto}.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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
> **Ojo — gemelos y cumplir años.** *Ser gemelos* — быть близнецами: {Los hijos de Rocío son gemelos.} Про двойняшек (непохожих) чаще говорят {mellizos}. Возраст «исполняется» глаголом {cumplir}: {cumplir 20 años}. Пока учим его как готовую фразу — {Kenji cumple 22 años el 8 de julio}, {Lara cumple 20 años en abril}; спряжение — *(Unidad 7)*.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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
> **Ojo — pareja y ex.** *La pareja* — и «пара» (двое), и «партнёр» любого пола: {mi pareja} — «мой партнёр / моя партнёрша», {tengo pareja} — «у меня есть отношения». *Ex* — самостоятельное слово: {mi ex} (бывший / бывшая), а также приставка: {mi ex novio, su ex marido, la ex mujer de mi tío}.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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
$SELECT Ejercicio 2 · ¿Soltero, casado o viudo?
ID: u05-s2-ej2-estado-civil
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INSTRUCTION: Выберите правильный вариант.
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: Lara no tiene marido: es _______.
OPTION: soltera | soltera
OPTION: casada | casada
OPTION: viuda | viuda
ANSWER: soltera
TEMPLATE: Los padres de Diego ya no son pareja: están _______.
OPTION: casados | casados
OPTION: divorciados | divorciados
OPTION: solteros | solteros
ANSWER: divorciados
TEMPLATE: El abuelo de Diego es _______: su mujer ya no está.
OPTION: soltero | soltero
OPTION: viudo | viudo
OPTION: casado | casado
ANSWER: viudo
TEMPLATE: El hermano de Diego tiene mujer: está _______.
OPTION: casado | casado
OPTION: separado | separado
OPTION: viudo | viudo
ANSWER: casado
TEMPLATE: Los hijos de Rocío tienen el mismo cumpleaños: son _______.
OPTION: adultos | adultos
OPTION: gemelos | gemelos
OPTION: ancianos | ancianos
ANSWER: gemelos
TEMPLATE: La hermana de Diego tiene catorce años: es _______.
OPTION: adolescente | adolescente
OPTION: adulta | adulta
OPTION: anciana | anciana
ANSWER: adolescente
TEMPLATE: Los gemelos tienen dos años: son _______ pequeños.
OPTION: niños | niños
OPTION: jóvenes | jóvenes
OPTION: mayores | mayores
ANSWER: niños
TEMPLATE: Sofía no está casada, pero tiene _______: su novio Luca.
OPTION: boda | boda
OPTION: pareja | pareja
OPTION: ex | ex
ANSWER: pareja
TEMPLATE: La madre de Diego está divorciada; su primer marido es ahora su _______.
OPTION: pareja | pareja
OPTION: ex | ex
OPTION: novio | novio
ANSWER: ex
TEMPLATE: El abuelo de Diego tiene ochenta y dos años: es una persona _______, pero muy activa.
OPTION: joven | joven
OPTION: mayor | mayor
OPTION: adolescente | adolescente
ANSWER: mayor
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$PRODUCE Ejercicio 2.1 · ¿Soltero, casado o viudo? (traducir)
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
ID: u05-s2-ej2-1-traducir
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INSTRUCTION: Переведите на испанский.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
INPUT: type
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
CHECK: llm
TEMPLATE: Я не замужем.
RESPONSE: Soy soltera.
TEMPLATE: Мой брат женат, и у него близнецы.
RESPONSE: Mi hermano está casado y tiene gemelos.
TEMPLATE: Родители Diego разведены.
RESPONSE: Los padres de Diego están divorciados.
TEMPLATE: Мой дедушка вдовец, ему восемьдесят два года.
RESPONSE: Mi abuelo es viudo, tiene ochenta y dos años.
TEMPLATE: У Sofía есть партнёр в Риме.
RESPONSE: Sofía tiene pareja en Roma.
TEMPLATE: Моя сестра — подросток: ей четырнадцать лет.
RESPONSE: Mi hermana es adolescente: tiene catorce años.
TEMPLATE: Близнецы — маленькие дети, им два года.
RESPONSE: Los gemelos son niños pequeños, tienen dos años.
TEMPLATE: Kenji исполняется двадцать два года восьмого июля.
RESPONSE: Kenji cumple veintidós años el 8 de julio.
TEMPLATE: Мои родители не разведены, но живут раздельно.
RESPONSE: Mis padres no están divorciados, pero están separados.
TEMPLATE: Бывший муж моей тёти женат на другой женщине; свадьба — в мае.
RESPONSE: El ex marido de mi tía está casado con otra mujer; la boda es en mayo.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
$SESSION Lección 3 · El carácter
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$DIALOGUE ¿Cómo es tu compañero?
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
ID: u05-s3-dialogo-caracter
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INTRO: На занятии профессор Elena даёт задание в парах: описать соседа по парте. В конце ребята описывают саму Elena.
INSTRUCTION: Слушайте и повторяйте каждую реплику.
REPEAT
VOCAB: el carácter
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: en parejas
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
Elena: Hoy trabajamos en parejas. La pregunta es: «¿Cómo es tu compañero o tu compañera?». Diego, ¿cómo es Lara?
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: agradable
VOCAB: abierto
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: abierta
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: sociable
VOCAB: inteligente
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: ser una persona agradable
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Diego: Lara es muy agradable y abierta. Es sociable: habla con todos. Y es muy inteligente.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: alegre
VOCAB: divertido
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: divertida
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: nervioso
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: nerviosa
VOCAB: tranquilo
VOCAB: tranquila
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Lara: ¡Gracias! Diego es alegre y muy divertido. Pero hoy está un poco nervioso: tiene examen de inglés.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Diego: No estoy nervioso, ¡estoy tranquilo! Bueno… un poco nervioso, sí.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Elena: Muy bien. ¿Y Kenji?
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: reservado
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: reservada
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: tímido
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: tímida
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: antipático
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: antipática
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: simpático
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: simpática
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: trabajador
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: trabajadora
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: serio
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: seria
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Sofia: Kenji es reservado y un poco tímido, pero no es antipático: es simpático y muy trabajador. En clase es serio y tranquilo.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: tolerante
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Kenji: Sofía es tolerante y alegre. Es una persona muy agradable. Y no es tímida: ¡es muy sociable!
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Elena: ¡Perfecto! Y ahora, otra pregunta: ¿cómo soy yo?
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Lara: Usted es simpática, abierta y muy tolerante con nosotros. Y no es seria: ¡es divertida!
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 Elena
ID: u05-s3-dialogo-caracter-role-elena
REFERENCE: u05-s3-dialogo-caracter
PLAY_ROLE: Elena
INTRO: Тот же разговор, но теперь Elena — это вы. Говорите её реплики в паузах.
INSTRUCTION: Вы — Elena. Произносите её реплики.
$GRAMMAR Léxico en contexto: pares de antónimos y ser / estar
Характер удобно учить парами противоположностей: одно слово тянет за собой второе. Все они — прилагательные, поэтому согласуются в роде и числе *(Unidad 4)*: {abierto → abierta → abiertos → abiertas}; слова на *-e* ({agradable, inteligente, tolerante, sociable, alegre}) по роду не меняются.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
| + | − |
|---|---|
| simpático/a | antipático/a |
| abierto/a | reservado/a |
| sociable | tímido/a |
| alegre | serio/a |
| tranquilo/a | nervioso/a |
| divertido/a | aburrido/a *(Unidad 1)* |
| trabajador/a | perezoso/a *(доп. вокабуляр)* |
| tolerante | intolerante |
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
Одно и то же прилагательное может описывать характер ({ser}) или состояние сейчас ({estar}) — повтор *(Unidad 1)*:
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
| ser + adjetivo (какой по характеру) | estar + adjetivo (какой сейчас) |
|---|---|
| {Diego es nervioso.} — он нервный человек | {Diego está nervioso.} — он нервничает сегодня |
| {Kenji es tranquilo.} — спокойный по натуре | {Kenji está tranquilo.} — сейчас спокоен |
| {Sofía es alegre.} — весёлая по характеру | {Sofía está alegre.} — у неё хорошее настроение |
| {Elena es seria.} — серьёзный человек | {Elena está seria.} — сейчас серьёзна |
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
Только с {ser}: {inteligente, trabajador/a, tolerante, sociable, tímido/a, agradable, simpático/a, divertido/a} — это черты, а не состояния.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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
> **Ojo — ser una persona agradable.** Когда прилагательное стоит после {persona}, оно согласуется с этим словом, то есть всегда в женском роде — даже о мужчине: {Diego es una persona muy agradable y abierta.} Без {persona} — согласование с человеком: {Diego es abierto.}
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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
> **Ojo — odiar, llevarse bien / mal con.** Эти глаголы — часть темы «характер», поэтому они в словаре, но спрягать их мы ещё не умеем: {odiar} — как обычный глагол на *-ar* *(Unidad 7)*, {llevarse bien con alguien} («ладить с кем-то») и {llevarse mal con alguien} («не ладить») — возвратный *(Unidad 8)*. Пока говорим о людях через {ser / estar + adjetivo}: {Es antipático} вместо «я его не выношу», {Somos amigos} вместо «мы ладим».
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$PRODUCE Ejercicio 3 · ¿Cómo es?
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
ID: u05-s3-ej3-como-es
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INSTRUCTION: Дополните прилагательное характера по подсказке — в правильной форме (род и число).
INPUT: type
CHECK: exact
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Lara es muy ____{sociable}: habla con todos. (общительная)
RESPONSE: Lara es muy sociable: habla con todos.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Kenji es un poco ____{tímido} y reservado. (застенчивый)
RESPONSE: Kenji es un poco tímido y reservado.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Diego es ____{alegre} y divertido. (весёлый — по характеру)
RESPONSE: Diego es alegre y divertido.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Sofía es una persona muy ____{agradable}. (приятная)
RESPONSE: Sofía es una persona muy agradable.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Hoy Diego está ____{nervioso}: tiene examen. (нервный)
RESPONSE: Hoy Diego está nervioso: tiene examen.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: La profesora Elena es ____{tolerante} con nosotros. (терпимая)
RESPONSE: La profesora Elena es tolerante con nosotros.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Kenji es muy ____{trabajador}: estudia y trabaja. (трудолюбивый)
RESPONSE: Kenji es muy trabajador: estudia y trabaja.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Mis compañeros de piso son ____{simpáticos} y abiertos. (приятные, милые)
RESPONSE: Mis compañeros de piso son simpáticos y abiertos.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Marina y Sofía no son ____{serias}: son muy ____{divertidas}. (серьёзные; весёлые, забавные)
RESPONSE: Marina y Sofía no son serias: son muy divertidas.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
TEMPLATE: Mi hermano es ____{tranquilo} por carácter, pero hoy está ____{alegre} porque hay una fiesta. (спокойный; весёлый, в хорошем настроении)
RESPONSE: Mi hermano es tranquilo por carácter, pero hoy está alegre porque hay una fiesta.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$PRODUCE Ejercicio 3.1 · ¿Cómo es? (traducir)
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
ID: u05-s3-ej3-1-traducir
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INSTRUCTION: Переведите на испанский.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
INPUT: type
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
CHECK: llm
TEMPLATE: Какой Diego? — Весёлый и очень забавный.
RESPONSE: ¿Cómo es Diego? — Es alegre y muy divertido.
TEMPLATE: Lara очень открытая и общительная.
RESPONSE: Lara es muy abierta y sociable.
TEMPLATE: Kenji сдержанный, но не неприятный.
RESPONSE: Kenji es reservado, pero no es antipático.
TEMPLATE: Elena — очень приятный человек.
RESPONSE: Elena es una persona muy agradable.
TEMPLATE: Я не нервный, я спокойный!
RESPONSE: ¡No soy nervioso, soy tranquilo!
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: ¡No soy nerviosa, soy tranquila!
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
TEMPLATE: Сегодня Diego нервничает: у него экзамен.
RESPONSE: Hoy Diego está nervioso: tiene examen.
TEMPLATE: Мои соседи по квартире умные и трудолюбивые.
RESPONSE: Mis compañeros de piso son inteligentes y trabajadores.
TEMPLATE: Sofía терпимая и весёлая.
RESPONSE: Sofía es tolerante y alegre.
TEMPLATE: Мой брат серьёзный и немного застенчивый, а моя сестра очень общительная.
RESPONSE: Mi hermano es serio y un poco tímido, y mi hermana es muy sociable.
TEMPLATE: Вы (usted) очень терпимы с нами и не серьёзны: Вы забавная!
RESPONSE: Usted es muy tolerante con nosotros y no es seria: ¡es divertida!
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
$SESSION Lección 4 · Celebraciones y gente alrededor
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$DIALOGUE La fiesta de cumpleaños de 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
ID: u05-s4-dialogo-cumpleanos
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INTRO: 8 июля — день рождения Kenji. Sofía и Lara устраивают в квартире праздник-сюрприз. Приходят Diego с Marina, две одногруппницы и соседка с третьего этажа.
INSTRUCTION: Слушайте и повторяйте каждую реплику.
REPEAT
VOCAB: ¡Felicidades!
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: la fiesta de cumpleaños
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: hacer una fiesta
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
VOCAB: ¡Sorpresa!
Sofia: ¡Felicidades, Kenji! ¡Veintidós años! Esta es tu fiesta de cumpleaños. ¡Sorpresa!
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: el invitado
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: la invitada
VOCAB: la invitación
VOCAB: invitar
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Kenji: ¡Muchas gracias! ¿Una fiesta para mí? ¿Y quiénes son los invitados?
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: el compañero de clase
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: la compañera de clase
VOCAB: el compañero de piso
VOCAB: la compañera de piso
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: el vecino
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: la vecina
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Lara: Tus compañeros de piso, claro. Diego y Marina, dos compañeras de clase y la vecina del tercero con su hija.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: el conocido
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: la conocida
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: el regalo
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: regalar
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Marina: Hola, Kenji. Soy Marina, la amiga de Diego. Tú y yo somos conocidos… pero hoy, amigos. Esto es un regalo para ti.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Kenji: ¡Gracias! ¿Qué es?
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: el socio
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: la socia
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Marina: Una invitación para el club de tenis: mi padre es socio del club. Es para dos personas.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: la tarta
VOCAB: la vela
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: de mi parte
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Diego: Y esta tarta es de mi parte: de chocolate, con veintidós velas. Ah, y una noticia: el examen de inglés… ¡muy bien!
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
VOCAB: ¡Enhorabuena!
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: felicitar
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Marina: ¡Enhorabuena, Diego! Y ahora, la foto: ¡todos con Kenji!
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Kenji: Gracias a todos. Sois unos compañeros de piso y unos amigos geniales.
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 Kenji
ID: u05-s4-dialogo-cumpleanos-role-kenji
REFERENCE: u05-s4-dialogo-cumpleanos
PLAY_ROLE: Kenji
INTRO: Тот же разговор, но теперь Kenji — это вы. Говорите его реплики в паузах.
INSTRUCTION: Вы — Kenji. Произносите его реплики.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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
$GRAMMAR Léxico en contexto: celebraciones y gente alrededor
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
Слова праздника ходят устойчивыми связками — учим их целиком, вместе с глаголом.
| Colocación | Traducción | Ejemplo |
|---|---|---|
| hacer una fiesta | устраивать вечеринку | {Hoy hay una fiesta en el piso.} |
| fiesta de cumpleaños | праздник в день рождения | {Es la fiesta de cumpleaños de Kenji.} |
| dar / hacer un regalo = regalar | дарить подарок | {Es un regalo para ti.} |
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
| invitar a alguien (a una fiesta) | приглашать кого-то | {la invitación, los invitados} |
| felicitar a alguien | поздравлять кого-то | {¡Felicidades! / ¡Enhorabuena!} |
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
| tarta de cumpleaños con velas | торт со свечами | {una tarta con veintidós velas} |
| cumplir 22 años | исполняться (22 года) | {Kenji cumple 22 años.} *(Unidad 7)* |
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
Люди вокруг нас: слово {compañero/a} само по себе значит «товарищ, напарник», а {de} + место уточняет, откуда мы друг друга знаем.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
| Español | Quién es |
|---|---|
| compañero de clase | учится с тобой в одной группе |
| compañero de piso | живёт с тобой в одной квартире *(Unidad 1)* |
| compañero de trabajo | работает с тобой |
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
| socio/a | член клуба ({socio del club de tenis}) или партнёр по бизнесу ({socios de una empresa}) |
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
| vecino/a | живёт рядом: в доме, на этаже, на улице |
| conocido/a | знакомый — уже не чужой, ещё не друг |
| amigo/a | друг |
| en parejas | по двое: {trabajar en parejas} — работать в парах (в классе) |
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
> **Ojo — ¡Felicidades! или ¡Enhorabuena!** {¡Felicidades!} — с днём рождения, с праздником, с Новым годом: поздравляем с датой. {¡Enhorabuena!} — с успехом: сданный экзамен, новая работа, свадьба, рождение ребёнка. На день рождения {¡Enhorabuena!} звучит странно, а на экзамен — {¡Felicidades!} можно, но реже.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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
> **Ojo — socio.** {Socio/a} — не «социальный» и не «сосед»: это член клуба или ассоциации (платит взнос, имеет карточку) либо партнёр по бизнесу. Сосед — {vecino/a}; сосед по квартире — {compañero/a de piso}.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$SELECT Ejercicio 4 · ¿Quién es quién?
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: u05-s4-ej4-quien-es-quien
INSTRUCTION: Сопоставьте начало фразы (1–10) с её концом (a–j).
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
OPTION: a | vecino/a
OPTION: b | compañero/a de trabajo
OPTION: c | invitado/a
OPTION: d | compañero/a de clase
OPTION: e | socio/a
OPTION: f | ¡Enhorabuena!
OPTION: g | en parejas
OPTION: h | conocido/a
OPTION: i | tarta de cumpleaños
OPTION: j | compañero/a de piso
TEMPLATE: Está en mi clase de español: es mi…
ANSWER: d
TEMPLATE: Vive en el piso con nosotros: es mi…
ANSWER: j
TEMPLATE: Trabaja en mi oficina: es mi…
ANSWER: b
TEMPLATE: Vive en el segundo, debajo de nosotros: es mi…
ANSWER: a
TEMPLATE: Tiene tarjeta del club de tenis: es…
ANSWER: e
TEMPLATE: Es amigo de un amigo: «hola» y «adiós», nada más. Es un…
ANSWER: h
TEMPLATE: Está en la fiesta porque tiene invitación: es…
ANSWER: c
TEMPLATE: Es de chocolate y tiene veintidós velas: es una…
ANSWER: i
TEMPLATE: «Diego, tu examen de inglés: ¡muy bien!» La palabra para este momento es…
ANSWER: f
TEMPLATE: Elena, en clase: «Diego con Lara, Sofía con Kenji». Trabajamos…
ANSWER: g
$PRODUCE Ejercicio 4.1 · ¿Quién es quién? (traducir)
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
ID: u05-s4-ej4-1-traducir
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INSTRUCTION: Переведите на испанский.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
INPUT: type
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
CHECK: llm
TEMPLATE: Поздравляю, Kenji! Это подарок для тебя.
RESPONSE: ¡Felicidades, Kenji! Es un regalo para ti.
TEMPLATE: Сегодня день рождения Kenji: ему исполняется двадцать два года.
RESPONSE: Hoy es el cumpleaños de Kenji: cumple veintidós años.
TEMPLATE: Кто гости?
RESPONSE: ¿Quiénes son los invitados?
TEMPLATE: Торт шоколадный, с двадцатью двумя свечами.
RESPONSE: La tarta es de chocolate, con veintidós velas.
TEMPLATE: Marina — подруга Diego; мы знакомые.
RESPONSE: Marina es la amiga de Diego; somos conocidos.
TEMPLATE: Мой отец — член теннисного клуба.
RESPONSE: Mi padre es socio del club de tenis.
TEMPLATE: Соседка с третьего этажа — тоже гостья.
RESPONSE: La vecina del tercero también es invitada.
TEMPLATE: В классе мы работаем в парах.
RESPONSE: En clase trabajamos en parejas.
TEMPLATE: В квартире вечеринка: приглашение — для одногруппников и соседей.
RESPONSE: En el piso hay una fiesta: la invitación es para los compañeros de clase y los vecinos.
TEMPLATE: Поздравляю (с успехом), Diego! Экзамен по английскому — очень хорошо! Это от меня.
RESPONSE: ¡Enhorabuena, Diego! El examen de inglés, ¡muy bien! Esto es de mi parte.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
$SESSION Lectura y comprensión
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$DIALOGUE Mensaje para 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: u05-s5-lectura
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INTRO: Вечером после праздника Lara пишет Valentina в Ciudad de México — та на этой же неделе заполняет formulario в своей школе. Прочитай сообщение.
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
Narrator: ¡Hola, Vale! Esta es mi primera semana en la academia.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
LINE: Ya tengo mi tarjeta de estudiante: en la secretaría hay un formulario con nombre, apellidos, fecha y lugar de nacimiento, dirección, código postal, número de teléfono y firma.
LINE: Estado civil: soltera, claro. ¿Tú también tienes formulario en tu escuela de México?
LINE: Mis compañeros de piso son geniales.
LINE: Diego es alegre y muy divertido, pero un poco nervioso por sus exámenes.
LINE: Su familia es grande: sus padres están divorciados, su hermano mayor está casado y tiene gemelos de dos años, y su abuelo es viudo.
LINE: Sofía es abierta y tolerante; Kenji es reservado y tímido, pero muy trabajador.
LINE: La profesora Elena es una persona muy agradable.
LINE: Hoy es el cumpleaños de Kenji: cumple veintidós años.
LINE: Hay una fiesta en el piso con tarta, velas y regalos.
LINE: Los invitados son los compañeros de clase, Marina —una amiga de Diego— y la vecina del tercero.
LINE: ¡Felicidades, Kenji! Un abrazo, Lara.
$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: u05-s5-comprension-opcion
REFERENCE: u05-s5-lectura
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
INSTRUCTION: Выберите правильный вариант.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
SHOW_PROMPT
PROMPT: En la secretaría de la academia hay…
OPTION: a | un formulario
OPTION: b | una fiesta
OPTION: c | un examen
ANSWER: a
PROMPT: El estado civil de Lara es…
OPTION: a | casada
OPTION: b | soltera
OPTION: c | divorciada
ANSWER: b
PROMPT: El hermano mayor de Diego…
OPTION: a | es viudo
OPTION: b | está divorciado
OPTION: c | está casado y tiene gemelos
ANSWER: c
PROMPT: Kenji es…
OPTION: a | antipático
OPTION: b | reservado y trabajador
OPTION: c | nervioso y abierto
ANSWER: b
PROMPT: Hoy Kenji cumple…
OPTION: a | veinte años
OPTION: b | treinta años
OPTION: c | veintidós años
ANSWER: c
$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: u05-s5-comprension-preguntas
REFERENCE: u05-s5-lectura
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INSTRUCTION: Ответьте на вопросы по тексту.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
INPUT: speak
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
CHECK: llm
SHOW_PROMPT
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
RUBRIC: Ключ — факт из текста; принять любую формулировку с верным фактом, полное предложение предпочтительно.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
PROMPT: ¿Qué datos hay en el formulario de la academia?
RESPONSE: Nombre, apellidos, fecha y lugar de nacimiento, dirección, código postal, número de teléfono y firma.
PROMPT: ¿Cómo es Diego?
RESPONSE: Es alegre y muy divertido, pero un poco nervioso por sus exámenes.
PROMPT: ¿Cómo es la profesora Elena?
RESPONSE: Es una persona muy agradable.
PROMPT: ¿Quiénes son los invitados a la fiesta?
RESPONSE: Los compañeros de clase, Marina y la vecina del tercero.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
$SESSION Lección 5 · Repaso y tarea final
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$PRODUCE Ejercicio 5 · Repaso de la Unidad 5
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: u05-s6-ej5-repaso
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INSTRUCTION: Переведите на испанский.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
INPUT: type
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
CHECK: llm
TEMPLATE: Как Ваша фамилия? — Моя фамилия García.
RESPONSE: ¿Cómo se apellida usted? — Me apellido García.
TEMPLATE: Дата рождения: двенадцатое апреля 2006 года.
RESPONSE: Fecha de nacimiento: el 12 de abril de 2006.
TEMPLATE: Ваша подпись здесь, пожалуйста. Вот Ваш студенческий билет.
RESPONSE: Su firma aquí, por favor. Esta es su tarjeta de estudiante.
TEMPLATE: Мой брат женат, и у него близнецы; мои родители разведены.
RESPONSE: Mi hermano está casado y tiene gemelos; mis padres están divorciados.
TEMPLATE: Мой дедушка вдовец. Он пожилой, но очень активный.
RESPONSE: Mi abuelo es viudo. Es mayor, pero muy activo.
TEMPLATE: Lara открытая и общительная; Kenji сдержанный и застенчивый.
RESPONSE: Lara es abierta y sociable; Kenji es reservado y tímido.
TEMPLATE: Elena — очень приятный человек.
RESPONSE: Elena es una persona muy agradable.
TEMPLATE: Поздравляю, Kenji! Это подарок для тебя.
RESPONSE: ¡Felicidades, Kenji! Es un regalo para ti.
TEMPLATE: Гости — одногруппники, соседи по квартире и соседка с третьего этажа.
RESPONSE: Los invitados son los compañeros de clase, los compañeros de piso y la vecina del tercero.
TEMPLATE: Kenji исполняется двадцать два года: есть торт с двадцатью двумя свечами, и в квартире вечеринка.
RESPONSE: Kenji cumple veintidós años: hay una tarta con veintidós velas y una fiesta en el piso.
$PRODUCE Tarea final · Tu ficha personal
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: u05-s6-tarea-final
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INTRO: Заполни свою анкету (ficha) — как в секретариате школы — и добавь к ней короткое «кто я».
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
INSTRUCTION: Заполните свою ficha по образцу и напишите 3–4 фразы о себе.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
INPUT: type
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
CHECK: llm
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
LONGFORM
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) все поля анкеты (nombre, apellidos, fecha y lugar de nacimiento, nacionalidad, sexo, estado civil, dirección, código postal, número de teléfono, documento, firma); 2) этап жизни и возраст; 3) семейное положение; 4) три черты характера; 5) одну строку о празднике. · FICHA DEL ESTUDIANTE · Nombre: ____ · Apellidos: ____ · Fecha de nacimiento: ____ · Lugar de nacimiento: ____ · Nacionalidad: ____ · Sexo: ____ · Estado civil: ____ · Dirección: ____ · Código postal: ____ · Número de teléfono: ____ · Documento: ____ · Firma: ____ · Soy ____.
RESPONSE: Nombre: Sofía · Apellidos: Ricci Conti · Fecha de nacimiento: 03/09/2001 · Lugar de nacimiento: Roma (Italia) · Nacionalidad: italiana · Sexo: mujer · Estado civil: soltera · Dirección: calle Betis, 14, 2.º, Sevilla · Código postal: 41010 · Número de teléfono: 611 22 33 44 · Documento: pasaporte YA1234567 · Firma: Sofía Ricci. Soy joven: tengo veinticuatro años. Soy soltera, pero tengo pareja. Soy una persona abierta, alegre y tolerante. Mi cumpleaños es el 3 de septiembre: ¡fiesta en el piso, con tarta, velas y muchos invitados!
RUBRIC: Принять любую корректно заполненную анкету и связный текст из 3-4 предложений с возрастом, семейным положением, характером и упоминанием дня рождения.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
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 Mediación · Ayuda a Valentina
ID: u05-s6-mediacion
INTRO: Твоя подруга Valentina только начала курс и пока говорит по-испански совсем немного. Secretaria спрашивает данные — ответь за неё.
INSTRUCTION: Ответьте за Valentina в 4–5 фразах: имя, дата и место рождения, семейное положение, номер телефона и одна черта характера.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
INPUT: type
CHECK: llm
LONGFORM
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: Ответьте за Valentina (имя, дата и место рождения, семейное положение, номер телефона, одна черта характера).
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
RESPONSE: Es mi amiga Valentina. Su fecha de nacimiento es el doce de marzo de 2005; su lugar de nacimiento es un secreto, pero está en el pasaporte. Es soltera. Su número de teléfono es el 622 11 22 33. Es una persona muy sociable y alegre.
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, передающий данные Valentina в третьем лице.
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$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: u05-s6-mini-test
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INSTRUCTION: Дополните или выберите вариант.
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
INPUT: type
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
CHECK: llm
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: ¿Cómo se _______ usted? — García López. (носить фамилию)
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
RESPONSE: apellida
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: Fecha de _______: 12/04/2006. (рождение)
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
RESPONSE: nacimiento
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: Mi hermano _______ casado y tiene gemelos. (ser / estar)
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
RESPONSE: está
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: Mi abuelo _______ viudo. (ser / estar)
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
RESPONSE: es
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: Kenji es reservado y un poco _______. (застенчивый)
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
RESPONSE: tímido
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: ¡_______, Kenji! Hoy es tu cumpleaños. (поздравление с днём рождения)
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
RESPONSE: Felicidades
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: Diego, Lara y Kenji viven en el piso de la calle Betis: son _______ de _______. (соседи по квартире)
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
RESPONSE: compañeros de piso
Regenerate units 0-5 in the current module dialect Adopts what the platform's parser now supports: \ instead of the deprecated \, a stable ID: on every activity, REFERENCE from the comprehension exercises to their reading passage, and inline gap answers (____{hay}) so each blank in a multi-gap item is graded on its own. Also replaces INPUT: either, which the parser silently downgrades to 'speak' — 37 exercises (translation drills, mini-tests, final tasks) were therefore refusing typed answers. They are now type or speak per exercise, with LONGFORM on the writing tasks. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 04:22:43 +03:00
$SESSION Vocabulario adicional
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$DIALOGUE Documentos
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: u05-s7-vocabulario-documentos
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
INTRO: Необязательные слова по теме — учите по желанию.
REPEAT
LINE: el visado
LINE: el permiso de residencia
LINE: el NIE
LINE: el certificado
LINE: el certificado de nacimiento
LINE: el carné de conducir
LINE: la tarjeta sanitaria
LINE: la fotocopia
LINE: la foto de carné
LINE: la cita previa
$DIALOGUE Más carácter
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: u05-s7-vocabulario-caracter
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
REPEAT
LINE: generoso
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: generosa
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
LINE: egoísta
LINE: optimista
LINE: pesimista
LINE: perezoso
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: perezosa
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
LINE: responsable
LINE: amable
LINE: paciente
LINE: cariñoso
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: cariñosa
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
LINE: honesto
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: honesta
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
$DIALOGUE La familia ampliada
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: u05-s7-vocabulario-familia
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
REPEAT
LINE: el suegro
LINE: la suegra
LINE: el cuñado
LINE: la cuñada
LINE: el sobrino
LINE: la sobrina
LINE: el nieto
LINE: la nieta
LINE: el padrino
LINE: la madrina
LINE: el yerno
LINE: la nuera
LINE: el bisabuelo
LINE: la bisabuela
LINE: los mellizos
LINE: las mellizas
LINE: el hijo único
LINE: la hija única
LINE: la familia numerosa
$DIALOGUE Verbos de la vida
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: u05-s7-vocabulario-verbos
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
REPEAT
LINE: nacer
LINE: crecer
LINE: cambiar
LINE: mudarse
LINE: enamorarse
LINE: casarse
LINE: tener hijos
LINE: jubilarse
LINE: jubilado
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: jubilada
add course manifest, 05.module and .gitignore
2026-09-22 03:14:56 +03:00
LINE: envejecer
LINE: morir
Reference in New Issue Copy Permalink