# Вайбкодинг для самых маленьких. Часть 1

DevFeed: [Вайбкодинг для самых маленьких. Часть 1](<https://devfeed.tech/articles/1-134445.md>)

Original publisher: [Read original article](<https://habr.com/ru/articles/1107966/>)

Author: Viacheslav-hub

Published: 2026-11-15T15:30:24Z

Content type: article

Language: ru

Sources: [Tagir Valeev](<https://devfeed.tech/sources/tagir-valeev.md>)

Topics: [codex](<https://devfeed.tech/topics/codex.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [Генеративные модели](<https://devfeed.tech/topics/tag-e493b9551702.md>), [Canvas](<https://devfeed.tech/topics/canvas.md>), [тест-кейсы](<https://devfeed.tech/topics/tag-532d1762d470.md>), [OpenAI](<https://devfeed.tech/topics/openai.md>), [Claude](<https://devfeed.tech/topics/claude.md>)

Tags: [acting](<https://devfeed.tech/tags/acting.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [anthropic](<https://devfeed.tech/tags/anthropic.md>), [anthropic-claude](<https://devfeed.tech/tags/anthropic-claude.md>), [api](<https://devfeed.tech/tags/api.md>), [canva](<https://devfeed.tech/tags/canva.md>), [canvas](<https://devfeed.tech/tags/canvas.md>), [chatgpt](<https://devfeed.tech/tags/chatgpt.md>), [chatgpt-plus](<https://devfeed.tech/tags/chatgpt-plus.md>), [claude](<https://devfeed.tech/tags/claude.md>), [codex](<https://devfeed.tech/tags/codex.md>), [llm](<https://devfeed.tech/tags/llm.md>), [tag-1605473766c5](<https://devfeed.tech/tags/tag-1605473766c5.md>), [tag-25b45a6c96e2](<https://devfeed.tech/tags/tag-25b45a6c96e2.md>), [tag-2919f56a7912](<https://devfeed.tech/tags/tag-2919f56a7912.md>), [tag-2c039dce53be](<https://devfeed.tech/tags/tag-2c039dce53be.md>), [tag-41e9f0e0da78](<https://devfeed.tech/tags/tag-41e9f0e0da78.md>), [tag-61cd5a476b1d](<https://devfeed.tech/tags/tag-61cd5a476b1d.md>), [tag-64251c106897](<https://devfeed.tech/tags/tag-64251c106897.md>), [tag-7ae2f4eba4d4](<https://devfeed.tech/tags/tag-7ae2f4eba4d4.md>), [tag-7c29fef020e9](<https://devfeed.tech/tags/tag-7c29fef020e9.md>), [tag-830ae51a4798](<https://devfeed.tech/tags/tag-830ae51a4798.md>), [tag-86b843454893](<https://devfeed.tech/tags/tag-86b843454893.md>), [tag-94b623af73a7](<https://devfeed.tech/tags/tag-94b623af73a7.md>), [tag-dc9376751d97](<https://devfeed.tech/tags/tag-dc9376751d97.md>), [tag-dcfd0059c9ac](<https://devfeed.tech/tags/tag-dcfd0059c9ac.md>)

## AI overview

The article explains how coding agents combine an LLM with tools and a Harness that manages context, state, and actions, then demonstrates a beginner workflow in Codex. It distinguishes Codex and Claude Code, which are tied to their respective model ecosystems; Cursor, which emphasizes editing alongside an agent and supports models from multiple developers; OpenCode, which is open source and allows provider or local model choice; and Hermes Agent, a broader assistant with persistent memory. The practical example builds Poster Studio in three natural-language requests: create a one-file HTML/CSS/JavaScript poster editor with Canvas PNG export, add styles and controls, then ask the agent to check specified behaviors. The author stresses that agent-reported checks do not guarantee correctness and recommends manually trying key scenarios. The approach is framed for personal, simple projects; complex work can bring misunderstandings, forgotten requirements, and regressions. Free Codex access is described as available with small limits, while stronger models and higher reasoning effort may cost more resources or time.

## Source excerpt

Что такое вайбкодинг и действительно ли для создания собственных приложений больше не нужно уметь программировать? В этой статье разберёмся, как работают ИИ-агенты, чем они отличаются от обычных языковых моделей и какие инструменты можно использовать для разработки. А затем на практике создадим собственное приложение с помощью Codex, не написав ни одной строчки кода. Читать далее