trae-community/trae-skills/cloudbase
Claude Code and agent skill by trae-community/trae-skills
- community
Checking that the SKILL.md file is still in the repository.
Use when building, deploying, or debugging Tencent CloudBase (腾讯云开发 / TCB) apps in Trae — Web, WeChat Mini Program, cloud functions, CloudRun, auth, NoSQL/PostgreSQL, storage, and built-in AI. Prefer CloudBase MCP tools first, then load published CloudBase skills for scenario details.
Install
npx vskill@latest install trae-community/trae-skills/cloudbaseSKILL.md
The same SKILL.md was published earlier as trae-community/trae-skills/cloudbase.
CloudBase (Tencent Cloud Development)
Description
Guide Trae to develop on Tencent CloudBase with the correct order of operations: environment binding → MCP management tools → scenario skill → implementation → review/deploy.
CloudBase covers Web / H5, WeChat Mini Program, cloud functions, CloudRun, document DB / PostgreSQL / MySQL, cloud storage, auth, and built-in AI models.
Usage Scenario
Use this skill when the user:
- Mentions CloudBase, 云开发, TCB, 腾讯云开发, or WeChat Mini Program cloud development
- Asks to scaffold, deploy, or debug a CloudBase Web / Mini Program / CloudRun app
- Needs auth, database, storage, cloud functions, or CloudBase AI model integration
- Wants Trae wired to
@cloudbase/cloudbase-mcp
Do not use for non-CloudBase backends or pure frontend work with no CloudBase dependency.
Preconditions
- Trae can run terminal commands and edit project files
- Prefer enabling CloudBase MCP (see below) before management API work
- User should have a Tencent Cloud / CloudBase account (browser login is prompted by MCP)
Instructions
- Confirm scenario — Web, WeChat Mini Program, CloudRun, database-only, auth-only, or AI model. State which scenario you will follow.
- Ensure CloudBase MCP is available in Trae
- Settings → MCP → Add → Manual:
- On first use, complete browser login / environment selection.
- Bind environment explicitly
- Call
envQuery(or equivalent) to resolve the canonicalEnvId. - Never pass aliases/nicknames directly into SDK init or deploy configs — resolve to full
EnvIdfirst.
- Prefer MCP tools for management work
- Auth providers, DB schema, functions, hosting, CloudRun, storage, security rules: use MCP first.
- Inspect tool schemas before calling; do not invent parameters.
- Load the matching published CloudBase skill before coding
- Main entry: https://github.com/TencentCloudBase/skills (or
npx skills add tencentcloudbase/cloudbase-skills) - Common ids:
web-development,miniprogram-development,auth-tool-cloudbase,auth-web-cloudbase,cloud-functions,cloudrun-development,postgresql-development-cloudbase,ui-design,cloudbase-platform - Read the skill fully before writing application code.
- Implementation order
- Resource prep via MCP (providers, collections/tables, permissions) → then frontend/backend code → local verify → deploy.
- Close-out
- Run a CloudBase-aware review when available (
cloudbase-code-reviewskill). - Report EnvId, resources touched, and preview/deploy URLs.
Examples
Example 1 — Web app with username/password auth
User: “用云开发做一个带登录的 Web 管理后台”
Agent should:
- Enable CloudBase MCP and bind
EnvId - Use auth MCP tools to enable username/password (and publishable key) before writing login UI
- Follow
auth-tool-cloudbasethenauth-web-cloudbase/web-development - Deploy static hosting only after auth providers are confirmed
Example 2 — WeChat Mini Program CRUD
User: “小程序云开发读写文档数据库”
Agent should:
- Confirm Mini Program + CloudBase scenario
- Use NoSQL MCP tools /
cloudbase-document-database-in-wechat-miniprogramguidance - Avoid Web SDK auth patterns in Mini Program code
References
- Docs: https://docs.cloudbase.net/ai/cloudbase-ai-toolkit/ai-agent-plugins
- Toolkit source: https://github.com/TencentCloudBase/CloudBase-AI-Toolkit
- Open Plugin / skills package: https://github.com/TencentCloudBase/cloudbase-plugin
- Published skills: https://github.com/TencentCloudBase/skills
Constraints
- Do not invent CloudBase API paths or MCP tool arguments
- Do not expose API keys /
service_rolecredentials in frontend code - After 2–3 failed attempts on the same path, stop and reroute (env, auth domain, permission model, or skill)
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/cloudbase/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/trae-community/trae-skills/cloudbase)