Ваша AI-идентичность и ваша AI-команда, которые работают на любой AI-платформе.
Войти Начать
Меню
Создать Agent of Me Обзор стилей Профессиональные агенты Агенты сообщества Рейтинг Новости об AI
AI-платформы Каталог Model Matrix Сравнить Какой AI мне использовать? Руководства по интеграции Настроить OpenClaw Prompt Fit
Обучение и инструменты Обучение Запрос к данным Конструктор агентов API
О сайте О нас Контакты Дисклеймеры
Войти Начать
Аккаунт
Ваша AI-идентичность, всегда с вами

Создайте бесплатный аккаунт и постройте свой профиль. По умолчанию всё приватно. Ничего не публикуется без вашего согласия.

Начать Войти
Тёмная тема

🧭 Режим с подсказками
Не знакомы с prompt, system instructions, context window, token? Мы объясняем каждый термин прямо при просмотре, простым языком. Те же страницы, со встроенными подсказками.

⚡ Экспертный режим
Вы уже знаете, как работает prompting. Только суть, чётко и компактно, без лишних пояснений. Это вид по умолчанию.

Язык интерфейса

Prompt Foundations · Раздел 4/5, Controlling the Output

Цели обучения
Нажмите «Далее» (или используйте стрелки), чтобы переходить по одной идее за раз. Таймера нет, тест из 5 вопросов ждёт в конце. Стрелка ← вверху выходит в любой момент; прогресс сохраняется.

Say the shape out loud

Models default to their house style: friendly paragraphs, medium length, often more than you wanted. Every part of that is controllable, but only if you say it: 'exactly three bullets', 'one paragraph, no list', 'a two-column table: risk, mitigation', 'subject line plus five sentences'.

Think about where the output is going. Text pasted into a slide, a ticket, a contract redline and a text message have different natural shapes. Name the destination and the shape follows.

Numbers hold better than adjectives

'Short' means nothing precise; '120 words or fewer' means one thing. 'A few options' invites five; 'exactly three, one sentence each' lands. When length or count matters, use numbers, and when a hard cap matters, say it is a hard cap.

The same goes for structure: 'start with the answer in one sentence, then give reasoning' beats 'be direct'. Behavior you can count is behavior you can specify.

Define done

The strongest format instruction is a definition of done: what must be true of the output for it to be finished? 'Done means: a customer can follow these steps without asking support anything.' 'Done means: every claim has a source or is labeled an assumption.'

Definitions of done catch what templates miss, because they describe quality, not just layout. They also give you a clean way to reject: 'not done, two claims are unlabeled'.

Templates: write it once, reuse forever

If you ask for the same shape weekly, stop describing it and start providing it: a template with headers and placeholders the model fills. 'Use exactly this structure: SITUATION / OPTIONS (max 3, with trade-offs) / RECOMMENDATION / RISKS.'

Templates are the bridge from prompting to systems: they turn taste into an artifact you can reuse, share and improve, the same idea a profile applies to your identity.

Мини-квиз, Controlling the Output

5 вопросов, каждый раз выбираются из банка случайно. Проходной балл 60%. Неограниченные попытки.

Следующий раздел: The Iteration Loop →
Читать полный текст урока

1. Say the shape out loud

Models default to their house style: friendly paragraphs, medium length, often more than you wanted. Every part of that is controllable, but only if you say it: 'exactly three bullets', 'one paragraph, no list', 'a two-column table: risk, mitigation', 'subject line plus five sentences'.

Think about where the output is going. Text pasted into a slide, a ticket, a contract redline and a text message have different natural shapes. Name the destination and the shape follows.

2. Numbers hold better than adjectives

'Short' means nothing precise; '120 words or fewer' means one thing. 'A few options' invites five; 'exactly three, one sentence each' lands. When length or count matters, use numbers, and when a hard cap matters, say it is a hard cap.

The same goes for structure: 'start with the answer in one sentence, then give reasoning' beats 'be direct'. Behavior you can count is behavior you can specify.

3. Define done

The strongest format instruction is a definition of done: what must be true of the output for it to be finished? 'Done means: a customer can follow these steps without asking support anything.' 'Done means: every claim has a source or is labeled an assumption.'

Definitions of done catch what templates miss, because they describe quality, not just layout. They also give you a clean way to reject: 'not done, two claims are unlabeled'.

4. Templates: write it once, reuse forever

If you ask for the same shape weekly, stop describing it and start providing it: a template with headers and placeholders the model fills. 'Use exactly this structure: SITUATION / OPTIONS (max 3, with trade-offs) / RECOMMENDATION / RISKS.'

Templates are the bridge from prompting to systems: they turn taste into an artifact you can reuse, share and improve, the same idea a profile applies to your identity.

Деловая активность

Business AnalystChief of StaffExecutive AssistantM&A AnalystManagement ConsultantOperations AnalystProject ManagerRecruiter

Финансы

AccountantDue Diligence AnalystEquity Research AnalystFamily Office AnalystFinancial AnalystFixed Income AnalystInvestment Banking AnalystPortfolio Analyst

Юридическое

Contract Review AssistantLegal Due Diligence AssistantLegal Research AssistantParalegal

Маркетинг

Brand StrategistContent StrategistGEO AnalystMarketing StrategistSEO AnalystSales Strategist

Личное

Career CoachLearning TutorReflection AssistantResearch AssistantTravel PlannerWriting Assistant

Недвижимость

Acquisition AnalystAsset Management AnalystCommercial Real Estate AnalystDevelopment AnalystLease AnalystProperty Financial Analyst

Исследование

Competitive Intelligence AnalystDeep Research AnalystIndustry Research AnalystJournalist ResearcherMarket Research AnalystMedical Research Assistant

Технологии

AI Strategy AdvisorCybersecurity Research AssistantData AnalystProduct ManagerSoftware Engineer