AI инструменты разработчика

AI оптимизатор Prompt

Делает из сырой идеи структурированный Prompt для кода исследования или продукта

БыстроМожно копироватьБез логина

Формат ввода

Например создать главную страницу AI tool сайта

Роль контекст задачи формат ограничения критерииСначала локальноПриватно по умолчанию

Input

Rough request

Use caseTemplate

Context

Результат

Optimized prompt

33 строк840 символов

Role
You are a senior full stack product engineer

Goal
Build an AI tool website with six practical developer tools

Context
Target users are beginners and indie developers. The page should be dense, useful and production ready.

Tasks
1. Clarify the smallest shippable version
2. Design the route data and component structure
3. Implement with existing project patterns
4. Add focused tests or manual verification
5. Explain exactly how to run it

Output Format
- Summary
- Assumptions
- Plan
- Deliverable
- Verification

Constraints
- Keep the MVP sharp
- Prefer existing stack and local helpers
- Avoid inflated marketing copy
- Ship something usable before adding breadth

Acceptance Criteria
- The feature works from the first screen
- Empty error loading and mobile states are handled
- The user can verify it with one command or URL

01Role

You are a senior full stack product engineer

02Goal

Build an AI tool website with six practical developer tools

03Context

Target users are beginners and indie developers. The page should be dense, useful and production ready.

04Tasks

1. Clarify the smallest shippable version 2. Design the route data and component structure 3. Implement with existing project patterns 4. Add focused tests or manual verification 5. Explain exactly how to run it

05Output Format

- Summary - Assumptions - Plan - Deliverable - Verification

06Constraints

- Keep the MVP sharp - Prefer existing stack and local helpers - Avoid inflated marketing copy - Ship something usable before adding breadth

07Acceptance Criteria

- The feature works from the first screen - Empty error loading and mobile states are handled - The user can verify it with one command or URL

Что делает

Prompt workflow

Converts a rough request into a structured promptСделано для быстрой копируемой работы

Adds role context task list output format and constraintsСделано для быстрой копируемой работы

Creates acceptance criteria so the answer is easier to verifyСделано для быстрой копируемой работы

Подходит для

Кому полезно

People writing with AI every day

Developers turning vague ideas into build tasks

Students making study or research prompts

Примеры

Ввод и вывод

Writing prompt optimization

Ввод: Write a product intro for a new AI tool.

Вывод: Audience, tone, structure, proof points, limits, and acceptance criteria before drafting.

Code generation prompt optimization

Ввод: Build a settings page in Next.js.

Вывод: Route, components, state, validation, error states, styling constraints, and verification commands.

Learning plan prompt optimization

Ввод: Help me learn Python in 30 days.

Вывод: Weekly goals, daily exercises, review checkpoints, final project, and measurable completion criteria.

FAQ и ограничения

Перед использованием

Does it call an AI model

No. The first version reshapes your request locally into a reusable prompt structure.

What makes a prompt stronger

A clear role context task list output format constraints and acceptance criteria.

Ограничения

Local template based output only
It does not guarantee model quality
Sensitive or private context should be removed before sharing with third party AI tools

Role You are a senior full stack product engineer Goal Build an AI tool website with six practical developer tools Context Target users are beginners and indie developers. The page should be dense, useful and production ready. Tasks 1. Clarify the smallest shippable version 2. Design the route data and component structure 3. Implement with existing project patterns 4. Add focused tests or manual verification 5. Explain exactly how to run it Output Format - Summary - Assumptions - Plan - Deliverable - Verification Constraints - Keep the MVP sharp - Prefer existing stack and local helpers - Avoid inflated marketing copy - Ship something usable before adding breadth Acceptance Criteria - The feature works from the first screen - Empty error loading and mobile states are handled - The user can verify it with one command or URL