Marky 网站构建器

从聊天提示到全栈 Web 应用

Marky 可以规划、编写、预览并发布网站;当创意需要真实应用逻辑时,还能构建后端、数据库和账户流程。

你的需求React + TypeScriptPostgres + 身份验证Cloud Run
AI SaaS 应用构建器示意图
AI 预约网站构建器示意图

直接回答

什么是 Marky AI 网站构建器?

Marky 是 Easy-Peasy.AI 中的对话式网站和 Web 应用构建器。它能生成可编辑的 React 与 TypeScript 项目,提供实时预览,添加 Express、Neon Postgres、身份验证和文件存储,并将批准的版本发布到 Google Cloud Run。你可以继续在聊天中迭代,或与 GitHub 同步代码。

选择合适的项目基础

简单网站还是全栈应用?

Marky can start light and add the application layer when the project needs accounts, stored records or protected server logic.

Comparison of simple website and full-stack Marky projects
功能简单网站全栈应用
FrontendVite, React, TypeScript and Tailwind CSSVite, React, TypeScript and Tailwind CSS
ServerNot includedExpress with typed tRPC APIs
DatabaseNot includedManaged Neon Postgres with Drizzle
User accountsNot includedBetter Auth foundation and Google OAuth support
File uploadsStatic project assetsCloudflare R2, using an S3-compatible API
Best forMarketing sites, portfolios and information pagesPortals, SaaS, directories and operational apps

生产级基础

对话背后的基础设施

The chat is the interface. Underneath it, Marky works with a conventional codebase and managed services you can reason about.

Preview
A public sandbox URL that updates while Marky works
Production
Google Cloud Run after you explicitly approve publishing
Data
A managed Neon Postgres branch for each full-stack project
Authentication
Better Auth foundation with centralized Google OAuth support
Domains
An editable subdomain, a connected domain or a purchased domain
Source code
Create or connect a GitHub repository and push project changes

完整功能图谱

Marky 能为整个 Web 项目完成什么

Capabilities are selected from the brief. A brochure site does not need a database; a portal does. Marky can add the deeper layers when the product actually calls for them.

The working code and feedback loop

Build and inspect

  • Vite, React, TypeScript and Tailwind CSS for every web project
  • Express and typed tRPC APIs when the project needs a backend
  • Live preview, project health checks, commands and tests for full-stack work
  • An upgrade path from a simple website to the full-stack scaffold

Persistent application data and identity

Store and protect

  • A managed Neon Postgres branch provisioned for the project
  • Drizzle models, schema changes, migrations and direct SQL operations
  • Better Auth foundations and centralized Google OAuth support
  • Cloudflare R2 object storage through its S3-compatible interface

Protected credentials and external capabilities

Connect services

  • Project secrets for private API keys and service credentials
  • Easy-Peasy.AI endpoints for text, chat, images, video, speech, transcription, bots and sounds
  • Third-party APIs for payments, email, calendars, maps and similar workflows
  • AWS APIs and services when configured—separate from the default Cloud Run publish path

Versions, domains and production operation

Ship and recover

  • Automatic preview checkpoints, named versions and rollback controls
  • Google Cloud Run publishing and an optional Always-On Hosting add-on
  • Editable subdomains plus custom-domain purchase or connection, DNS and SSL setup
  • GitHub repository creation, linking, push and later code synchronization

构建循环

从想法到生产环境的清晰路径

  1. 01

    Describe the outcome

    Give Marky the users, core workflow, records and visual direction—not just a page count.

  2. 02

    Review the live preview

    Marky scaffolds the project, runs it and opens a URL you can inspect while the conversation continues.

  3. 03

    Refine and checkpoint

    Request changes in plain language and save a named working version before larger revisions.

  4. 04

    Publish when it is ready

    Approve production deployment, choose the URL and optionally connect the codebase to GitHub.

明确边界

上线前需要了解的事项

  • Google Cloud Run is the built-in deployment target. AWS is possible by adapting the GitHub codebase or integrating AWS services with credentials.
  • Payments, email, maps, calendars and other third-party services require suitable APIs, your credentials and real integration testing.
  • Marky can implement metadata, structured data, sitemaps and internal links when requested; useful source content still determines search quality.
  • Apps handling regulated or highly sensitive data need an independent security, privacy and compliance review before production use.

常见问题

Marky 网站构建器常见问题

Can Marky build a full-stack website, not just a landing page?

Yes. Marky can start with a lightweight React website or create a full-stack web app with an Express backend, typed APIs, a Neon Postgres database, Drizzle models, authentication, file storage, and server-side integrations. The scaffold is chosen from the requirements in your brief.

Can a Marky website include Google login?

Yes. Full-stack projects include an authentication foundation and support a centralized Google OAuth flow. Marky can build the sign-in screens and authenticated product experience, then test the flow in preview before the app is published.

Where are Marky websites hosted?

The built-in production publishing flow deploys websites and web apps to Google Cloud Run. Every published project can use an editable easy-peasy.site subdomain, connect an existing custom domain, or purchase a new domain through the project settings.

Can Marky deploy a website to AWS?

Google Cloud Run is Marky's built-in deployment target. Because the generated project is a real TypeScript and Node.js codebase that can be pushed to GitHub, a developer can adapt it for AWS. AWS services and credentials are an integration path, not the default one-click publish flow.

Do I own or export the website code?

You can connect GitHub, create a new repository or link an existing one, and push the project code and later changes. Marky also keeps preview checkpoints so you can inspect versions and roll the editable project—or an eligible live deployment—back to an earlier state.

Can I add a database after starting with a simple website?

Yes. Marky can upgrade a static web project to the full-stack scaffold when the site grows to need user accounts, stored records, backend APIs, uploads, or AI features. The upgrade provisions the database and authentication foundation without requiring a brand-new conversation.

从清晰的需求说明开始

告诉 Marky 网站面向谁,以及用户需要完成什么。

开始构建