Build and deploy a full-stack app with CloudBase
One paragraph of requirements, one app you can put online: schema, row-level permissions, pages, and the deploy all come from the agent. You sign in and confirm.
The backend lives in one environment: PostgreSQL, authentication, object storage, cloud functions, static hosting. The frontend reads and writes directly, with no API layer to build yourself.
Frontend and backend both run here: Vite, Next.js, Nuxt, and Astro on the frontend; Express, NestJS, Spring, Gin, and FastAPI on the backend. Web, mini program, Flutter, and C# each have an SDK, and any other language goes through the HTTP API. Pure static builds ship to static hosting; SSR or a server of your own goes to CloudRun.
A database agents handle well
PostgreSQL keeps its rules in standard SQL — tables, constraints, functions, policies. An agent that reads the schema knows how data is stored, who can reach it, and what it is allowed to change, all in one pass instead of a dozen console clicks.
Full-text search, scheduled jobs, and vector search also run inside the database through functions and pgvector, so each one does not need a service of its own.
Connect
Copy this into your AI tool and it will wire CloudBase up:
Read https://docs.cloudbase.net/skill.md and connect me to CloudBase (China site).
Tell me when you are done, and suggest the most relevant next step.
Then just describe the app — no console clicks:
Build me a team todo board on CloudBase: email password sign-in, todos filtered by person,
create / complete / delete, avatar upload.
PostgreSQL for the database, a todos table with user_id, RLS enabled, all four policies scoped to auth.uid().
Frontend in Vite + React + TypeScript, reading and writing directly; cloud functions only for sending email.
Deploy to static hosting and give me the URL.
Prefer to set it up yourself? Pick your AI tool:
This client ships CloudBase built in. Setup follows its own marketplace or connectors, and the site follows your login state — nothing to pick here.
Configure your AI tool to connect with CloudBase capabilities. Supports local and remote connection. See connection modes.
Step 1: Install / Configure CloudBase
Open connector:
Open in WorkBuddyUse the built-in connector
1. In the WorkBuddy conversation interface, click the Connect App button at the bottom left of the input box 2. Find Tencent CloudBase in the popup, click to enter the detail page, and click Connect to enable
Once enabled, you can use CloudBase capabilities directly in the conversation.
Use the slash command
Type `/cloudbase` in a conversation to pick the CloudBase Skill, then describe what you need.
Bundled experts
WorkBuddy expert center ships four CloudBase experts; Full-Stack Engineer covers the most ground, and the other three are listed in the WorkBuddy setup guide.
Step 2: Chat with AI
Enter the following in your AI chat in order:
Read https://docs.cloudbase.net/skill.md and tell me what it can doUse CloudBase Skills: 帮我把 IDE 接到 CloudBase 远程 MCP，国内站和国际站分别用哪个地址？Make the agent fluent in CloudBase
The WorkBuddy Expert Center has a CloudBase full-stack expert: how to split tables, how to write policies, when to reach for a cloud function — for the app you are building, so you are not re-explaining the setup every time.
See CloudBase experts on WorkBuddy.
Documentation index
| What you need | Where to read |
|---|---|
| Create an environment, get a Publishable Key | Create a CloudBase environment |
| How to connect, per AI tool | Connect CloudBase MCP |
| Tables and row-level policies in full | PostgreSQL base permissions |
| Sign-in methods | Authentication |
| Uploads and access control | PG-mode storage |
| Cloud functions | Cloud functions |
| SSR and containerized deploys | CloudRun |
| Going live, binding a domain | Static website hosting |
| Node services, other languages | Node SDK, HTTP API |
| Mobile and desktop | Flutter SDK, C# SDK |
| Videos and hands-on cases | Videos and tutorials |
| Why PG mode is designed this way | The database agents love is here |