2026-09-22 03:55:56 +03:00
$MODULE Unidad 0 — El alfabeto, los sonidos y la lectura
2026-09-22 00:09:22 +00:00
FORMAT: 2
DIOCO_DOC_ID: lc_es_verano_sevilla_u00
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: Алфавит, звуки, правила чтения, ударение, вопросительные и восклицательные знаки.
2026-09-22 00:09:22 +00:00
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: Narrator | Kore | Narrador neutro
2026-09-22 04:22:43 +03:00
$SESSION Unidad 0
2026-09-22 00:09:22 +00:00
$GRAMMAR Objetivos y personajes
**Цели юнита**
- Узнать испанский алфавит и названия букв.
- Научиться читать по-испански: «как пишется, так и читается».
- Запомнить буквы-ловушки (c, g, j, h, ll, ñ, qu, r/rr, v, z).
- Понять, куда падает ударение, и зачем нужны ¿ ? ¡ !
- Прочитать вслух первый связный текст («Verano en Sevilla») под аудио.
**Хорошая новость:** испанский — почти фонетический язык. Выучив несколько правил, вы сможете прочитать почти любое слово, даже незнакомое.
2026-09-22 04:22:43 +03:00
$SESSION Lección 1 · El abecedario
2026-09-22 00:09:22 +00:00
$GRAMMAR El abecedario
2026-09-23 01:29:37 +03:00
В с У с
2026-09-22 00:09:22 +00:00
> **Что в квадратных скобках.** Это знаки IPA — международного фонетического алфавита. Они одинаковы для всех языков, поэтому показывают звук точно, без «примерно как буква в твоём языке». Знаки, которые встретятся чаще всего, разобраны сразу после таблицы.
2026-09-23 01:29:37 +03:00
| Буква | Название | Примеры | Звук (IPA) |
2026-09-22 00:09:22 +00:00
|---|---|---|---|
2026-09-23 01:29:37 +03:00
| A a | {a} | {amigo}, {casa}, {agua}, {mañana} | [a] |
| B b | {be} | {bueno}, {bien}, {bonito}, {bailar} | [b] |
| C c | {ce} | {casa}, {cosa} / {cine}, {cena} | [k] / [θ] |
| D d | {de} | {día}, {dos}, {dinero}, {domingo} | [d] |
| E e | {e} | {español}, {este}, {tener}, {café} | [e] |
| F f | {efe} | {foto}, {favor}, {fácil}, {café} | [f] |
| G g | {ge} | {gato}, {gracias} / {gente}, {gimnasio} | [g] / [x] |
| H h | {hache} | {hola}, {hora}, {hospital}, {hermano} | — (не читается) |
| I i | {i} | {isla}, {mi}, {libro}, {vivir} | [i] |
| J j | {jota} | {jamón}, {joven}, {trabajo}, {hijo} | [x] |
| K k | {ka} | {kilo}, {kiwi}, {koala}, {kárate} | [k] |
| L l | {ele} | {luna}, {hola}, {sol}, {plaza} | [l] |
| M m | {eme} | {madre}, {mesa}, {mano}, {amigo} | [m] |
| N n | {ene} | {noche}, {nada}, {bueno}, {tren} | [n] |
| Ñ ñ | {eñe} | {niño}, {España}, {año}, {mañana} | [ɲ] |
| O o | {o} | {ojo}, {hola}, {todo}, {como} | [o] |
| P p | {pe} | {padre}, {piso}, {plaza}, {papel} | [p] |
| Q q | {cu} | {queso}, {quién}, {aquí}, {parque} | [k] |
| R r | {erre} | {pero}, {cara} / {rosa}, {perro} | [ɾ] / [r] |
| S s | {ese} | {sol}, {casa}, {mesa}, {gracias} | [s] |
| T t | {te} | {tarde}, {todo}, {taxi}, {teléfono} | [t] |
| U u | {u} | {uno}, {luna}, {mucho}, {usted} | [u] |
| V v | {uve} | {vino}, {vivir}, {verde}, {viernes} | [b] (как B!) |
| W w | {uve doble} | {web}, {wifi}, {sándwich}, {waterpolo} | [w] / [b] (в заимств.) |
| X x | {equis} | {taxi}, {examen}, {éxito}, {extra} | [ks] |
| Y y | {ye} ({i griega}) | {yo}, {ya} / {y}, {hay} | [ʝ] / [i] |
| Z z | {zeta} | {zapato}, {zumo}, {plaza}, {azul} | [θ] |
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 Los símbolos IPA (как устроены эти звуки)
2026-09-22 00:09:22 +00:00
| Знак | Что делает рот | Пример |
|---|---|---|
2026-09-23 01:29:37 +03:00
| [θ] | кончик языка слегка высовывается между зубами, воздух проходит с
| [s] | кончик языка у
| [x] | язык поднят к задней части нёба, сильный шумный выдох — как при откашливании | {jamón}, {gente} |
| [ɲ] | кончик языка у а
| [ʝ] | средняя часть языка почти касается нёба, воздух проходит с
| [ɾ] | один быстрый удар кончика языка по нёбу за верхними зубами | {pero}, {cara} |
| [r] | тот же удар, но повторённый несколько раз подряд — язык вибрирует | {perro}, {Roma} |
| [b] | губы смыкаются; в испанском так читаются **и** *b*, **и** *v* | {vino}, {bueno} |
| [ˈ а ˈ
2026-09-22 00:09:22 +00:00
> **Nota — [θ] или [s]?** В В е г о О б а
$PRODUCE Ejercicio 1 · Las letras
2026-09-22 04:22:43 +03:00
ID: u00-s1-ej1-letras
2026-09-22 00:09:22 +00:00
INSTRUCTION: Напишите название буквы по-испански.
INPUT: type
CHECK: exact
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Ñ → ____{eñe}
2026-09-22 00:09:22 +00:00
RESPONSE: eñe
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: J → ____{jota}
2026-09-22 00:09:22 +00:00
RESPONSE: jota
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: H → ____{hache}
2026-09-22 00:09:22 +00:00
RESPONSE: hache
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: R → ____{erre}
2026-09-22 00:09:22 +00:00
RESPONSE: erre
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: V → ____{uve}
2026-09-22 00:09:22 +00:00
RESPONSE: uve
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: Y → ____{ye | i griega}
2026-09-22 00:09:22 +00:00
RESPONSE: ye
2026-09-22 04:22:43 +03:00
ACCEPT: i griega
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
TEMPLATE: Z → ____{zeta}
2026-09-22 00:09:22 +00:00
RESPONSE: zeta
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: G → ____{ge}
2026-09-22 00:09:22 +00:00
RESPONSE: ge
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: W → ____{uve doble}
2026-09-22 00:09:22 +00:00
RESPONSE: uve doble
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: Q → ____{cu}
2026-09-22 00:09:22 +00:00
RESPONSE: cu
2026-09-22 04:22:43 +03:00
$SESSION Lección 2 · Las vocales y las consonantes difíciles
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
$GRAMMAR Las 5 vocales
2026-09-22 00:09:22 +00:00
В В о
| Vocal | Звук | Рот | Пример |
|---|---|---|---|
2026-09-23 01:29:37 +03:00
| a | [a] | рот широко открыт, язык внизу | {casa} [ˈ
| e | [e] | рот приоткрыт, губы растянуты | {mesa} [ˈ
| i | [i] | губы сильно растянуты, язык высоко впереди | {día} [ˈ
| o | [o] | губы округлены | {hola} [ˈ
| u | [u] | губы вытянуты трубочкой | {uno} [ˈ
2026-09-22 00:09:22 +00:00
2026-09-23 01:29:37 +03:00
> Гласные не «глотаются» и не сокращаются: {Sevilla} = [se.ˈ
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
$GRAMMAR Diptongos y triptongos
2026-09-22 00:09:22 +00:00
Гласные делятся на **сильные** (a, e, o) и **слабые** (i, u).
- **Дифтонг** — две гласные в одном слоге, если хотя бы одна слабая (i/u). Слабая произносится коротко и сливается с о б е
| Дифтонг | Пример | Звучит как |
|---|---|---|
2026-09-23 01:29:37 +03:00
| ie | {bien}, {tiene} | [bjen] |
| ue | {bueno}, {agua} | [ˈ
| ai / ay | {aire}, {hay} | [ˈ
| ei / ey | {seis}, {rey} | [sei̯s] |
| au | {autobús} | [au̯.to.ˈ
| eu | {Europa} | [eu̯.ˈ
| oi / oy | {hoy}, {soy} | [oi̯] |
| iu / ui | {ciudad}, {muy} | [θju.ˈ
- **Трифтонг** — слабая + сильная + слабая в одном слоге: {estudiáis}, {Paraguay}, {buey}.
- **Две сильные гласные (a, e, o) Н Е
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 Consonantes-ловушки
2026-09-22 00:09:22 +00:00
Эти буквы читаются не так, как можно предположить по их виду:
| Буква / сочетание | Правило | Пример | Звучит как |
|---|---|---|---|
2026-09-23 01:29:37 +03:00
| **c + a/o/u** | [k] | {casa}, {cosa} | [ˈ ˈ
| **c + e/i** | [θ] (в Лат. Америке [s]) | {cena}, {cine} | [ˈ ˈ
| **z** | [θ] (в Лат. Америке [s]) | {zapato} | [θa.ˈ
| **g + a/o/u** | [g] | {gato} | [ˈ
| **g + e/i** | [x] | {gente}, {gimnasio} | [ˈ
| **gue / gui** | [ge] / [gi], u немая | {guitarra} | [gi.ˈ
| **güe / güi** | u **читается**: [gwe] / [gwi] | {pingüino} | [pin.ˈ
| **j** | [x] | {jamón} | [xa.ˈ
| **h** | **не читается** | {hola}, {hora} | [ˈ ˈ
| **ll** | [ʝ] | {calle}, {llave} | [ˈ
| **ñ** | [ɲ] | {España}, {niño} | [es.ˈ
| **qu + e/i** | [k], u немая | {queso}, {quién} | [ˈ
| **r** | одиночный — [ɾ] | {pero} | [ˈ
| **rr / r-** в начале | раскатистый [r] | {perro}, {Roma} | [ˈ ˈ
| **v** | [b] (= B) | {vino}, {Sevilla} | [ˈ
| **x** | [ks] | {taxi}, {examen} | [ˈ
2026-09-22 00:09:22 +00:00
$SELECT Ejercicio 2 · ¿Cómo se lee?
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: u00-s2-ej2-como-se-lee
2026-09-22 04:22:43 +03:00
INSTRUCTION: Прочитайте вслух и отметьте, как звучит выделенная буква.
2026-09-22 00:09:22 +00:00
SHOW_PROMPT
OPTION: k | [k]
OPTION: θ | [θ]
OPTION: g | [g]
OPTION: gw | [gw]
OPTION: x | [x]
OPTION: ʝ | [ʝ]
OPTION: b | [b]
OPTION: r | [r]
PROMPT: gente
ANSWER: x
PROMPT: casa
ANSWER: k
PROMPT: jamón
ANSWER: x
PROMPT: calle
ANSWER: ʝ
PROMPT: vino
ANSWER: b
PROMPT: cine
ANSWER: θ
PROMPT: guitarra
ANSWER: g
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
FEEDBACK: u немая
2026-09-22 00:09:22 +00:00
PROMPT: zapato
ANSWER: θ
PROMPT: pingüino
ANSWER: gw
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
FEEDBACK: u читается
2026-09-22 00:09:22 +00:00
PROMPT: perro
ANSWER: r
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
FEEDBACK: раскатистый
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
$SESSION Lección 3 · Las reglas de lectura
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
$GRAMMAR Las 6 reglas básicas
2026-09-22 00:09:22 +00:00
Главное правило: **в испанском читается почти всё, как написано.** Нет «немых» букв, кроме **h**, и нет редукции гласных. Запомните несколько правил — и сможете прочитать любое слово.
2026-09-23 01:29:37 +03:00
1. **Каждая гласная читается всегда одинаково.** Безударные гласные не ослабляются: {teléfono} = [te.ˈ о б е
2. **H никогда не читается.** {hola} = [ˈ ˈ
2026-09-22 00:09:22 +00:00
3. **C и G меняют звук перед e/i.** Перед a/o/u — [k] и [g]; перед e/i — [θ] и [x].
2026-09-23 01:29:37 +03:00
4. **U немая в gue, gui, que, qui.** Она нужна, чтобы сохранить твёрдый звук: {guitarra}, {queso}. Если над ней стоят точки (**ü**), её читают: {pingüino}.
5. **V = B.** Одинаковый звук. {vaca} и {baca} звучат одинаково.
6. **R в начале слова и RR — раскатистые.** {Roma}, {perro} — вибрирующее [r]; {pero} — один удар языка, [ɾ].
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
$GRAMMAR El acento (ударение)
2026-09-22 00:09:22 +00:00
Без письменного знака ударение ставится так:
2026-09-23 01:29:37 +03:00
- Слово кончается на **гласную, -n или -s** → ударение на **предпоследнем** слоге: {casa} (**ca**-sa), {joven} (**jo**-ven), {libros} (**li**-bros).
- Слово кончается на **другую согласную** → ударение на **последнем** слоге: {hotel} (ho-**tel**), {ciudad} (ciu-**dad**), {español} (es-pa-**ñol**).
- **Tilde (´
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
$GRAMMAR Los signos ¿ ? ¡ !
2026-09-22 00:09:22 +00:00
В с
2026-09-23 01:29:37 +03:00
- {¿Cómo te llamas?} — Как тебя зовут?
- {¡Hola!} — Привет!
- {¿De dónde eres?} — Откуда ты?
2026-09-22 00:09:22 +00:00
$SELECT Ejercicio 3 · El acento
2026-09-22 04:22:43 +03:00
ID: u00-s3-ej3-acento
2026-09-22 00:09:22 +00:00
INSTRUCTION: Где ударение — на предпоследнем (ПП) или последнем (П) слоге? (без учёта tilde)
SHOW_PROMPT
OPTION: pp | Предпоследний слог (ПП)
OPTION: p | Последний слог (П)
PROMPT: casa
ANSWER: pp
FEEDBACK: ca-sa
PROMPT: hotel
ANSWER: p
FEEDBACK: ho-tel
PROMPT: español
ANSWER: p
FEEDBACK: es-pa-ñol
PROMPT: joven
ANSWER: pp
FEEDBACK: jo-ven
PROMPT: ciudad
ANSWER: p
FEEDBACK: ciu-dad
PROMPT: libros
ANSWER: pp
FEEDBACK: li-bros
PROMPT: hospital
ANSWER: p
FEEDBACK: hos-pi-tal
PROMPT: amigo
ANSWER: pp
FEEDBACK: a-mi-go
PROMPT: reloj
ANSWER: p
FEEDBACK: re-loj
PROMPT: jueves
ANSWER: pp
FEEDBACK: jue-ves
2026-09-22 04:22:43 +03:00
$SESSION Lección 4 · Lectura y repaso
2026-09-22 00:09:22 +00:00
$DIALOGUE Lee en voz alta
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: u00-s4-lee-en-voz-alta
2026-09-22 04:22:43 +03:00
INSTRUCTION: Прочитайте вслух. В с е
REPEAT
2026-09-22 00:09:22 +00:00
LINE: hola
LINE: gracias
LINE: España
LINE: Sevilla
LINE: español
LINE: amigo
LINE: casa
LINE: queso
LINE: gente
LINE: jamón
LINE: guitarra
LINE: ciudad
LINE: café
LINE: niño
LINE: calle
LINE: vino
LINE: taxi
LINE: hospital
LINE: teléfono
LINE: buenos días
$DIALOGUE Lectura · «Verano en Sevilla»
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: u00-s4-lectura
INTRO: Первый связный текст. Н е
INSTRUCTION: Слушайте аудио и читайте вслед за диктором.
2026-09-22 00:09:22 +00:00
Narrator: Lara está en Sevilla, en España.
LINE: Hola. Buenos días.
LINE: Lara es amiga de Kenji. Kenji es de Japón.
LINE: Por la mañana hay sol. En la calle hay gente: un niño, un taxi, un café.
LINE: En el café hay queso, jamón y vino. Kenji toca la guitarra.
LINE: Lara ve la ciudad: una plaza, una casa, un hospital. Gracias, Sevilla.
LINE: Por la noche Lara camina por la calle. Hay luna. Hay una guitarra. Lara está feliz.
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
NOTES: Как читать: сначала послушайте аудио целиком. Потом читайте по абзацам вслед за диктором. Следите за «ловушками»: gente [ˈ ˈ ˈ ˈ ˈ ˈ ˈ ˈ ˈ ˈ
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
$PRODUCE Ejercicio 4 · Deletrea
ID: u00-s4-ej4-deletrea
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
INSTRUCTION: Назовите буквы по-испански, как при диктовке имени.
2026-09-22 04:22:43 +03:00
INPUT: speak
CHECK: reveal
2026-09-22 00:09:22 +00:00
Publish the full 40-unit course in the current module dialect
Adds units 6-39 and regenerates 0-5, so the whole course now uses what the
platform's parser actually supports: \, a stable ID: per activity,
REFERENCE from comprehension exercises to their reading passage, inline gap
answers, and a real INPUT mode per exercise instead of 'either' (which the
parser silently turned into speak-only).
Each conversational dialogue is followed by a role-play replay: the learner
performs one character's lines from meaning while the script is blurred, and
the role alternates between Lara and her interlocutor down the unit.
The two revision units could not be converted in one pass — the model stops
producing after roughly a dozen activities, and the tail of the unit went
missing without any error. They are built from three-lesson chunks and
stitched, which is why 18 and 39 carry 65 and 80 activities.
1367 activities across 40 modules: the production parser accepts all of them,
the linter is silent, and the fidelity checker reports no missing vocabulary
row, dialogue line, exercise item or answer key.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-09-22 06:46:26 +03:00
EXAMPLE
TEMPLATE: Lara
RESPONSE: ele · a · erre · a
2026-09-22 04:22:43 +03:00
TEMPLATE: hola
RESPONSE: hache · o · ele · a
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: Kenji
RESPONSE: ka · e · ene · jota · i
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: Sevilla
RESPONSE: ese · e · uve · i · ele · ele · a
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: gracias
RESPONSE: ge · erre · a · ce · i · a · ese
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: queso
RESPONSE: cu · u · e · ese · o
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: niño
RESPONSE: ene · i · eñe · o
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: jamón
RESPONSE: jota · a · eme · o · ene
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: taxi
RESPONSE: te · a · equis · i
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: España
RESPONSE: e · ese · pe · a · eñe · a
2026-09-22 00:09:22 +00:00
2026-09-22 04:22:43 +03:00
TEMPLATE: guitarra
RESPONSE: ge · u · i · te · a · erre · erre · a
