MIT-licensed Tencent CloudBase toolkit — Plugin / Skills / MCP layer that lets AI coding tools manage database, functions, auth and deploys from chat; connects to DSH via standard MCP config (DeepSeek Harness not explicitly documented).
DSH integration
Compatible
Inferred
Safety audit
Unaudited
Last verified
2026-08-21
License
MIT
01What can it help you accomplish?
Build and ship backend resources — database, auth, functions, hosting — from an AI coding conversation
CloudBase resources created in your own environment: PostgreSQL / document DB data models and CRUD, cloud functions / Cloud Run, auth and storage, static hosting / Mini Program releases
Developers using MCP-capable AI coding tools who want Tencent CloudBase backend work done in chat instead of the cloud console
Operate and debug a live CloudBase backend without leaving the agent conversation
Login, data queries and changes, cloud function and hosting management, and log inspection executed through the MCP server
Operators and indie builders maintaining CloudBase apps who want data CRUD, deploys and log reads handled by the agent
02How to install into DeepSeek Harness
Prerequisites
- Node.js v18.15.0+
- A CloudBase environment
- An AI tool that supports Plugin / Skills / MCP
Installation steps
- 01
Install as a Plugin (when the tool supports it): `npx plugins add TencentCloudBase/cloudbase-plugin`
$ npx plugins add TencentCloudBase/cloudbase-plugin
- 02
Or install the CloudBase AI CLI: `npm i -g @cloudbase/cli && tcb ai`
$ npm i -g @cloudbase/cli && tcb ai
- 03
Or configure MCP manually in the IDE config file, e.g. `{"mcpServers":{"cloudbase":{"command":"npx","args":["@cloudbase/cloudbase-mcp@latest"]}}}`
Verify the integration
Not specified by the author
03DSH integration and capability boundaries
README does not mention DeepSeek Harness explicitly; usable from DSH through the generic MCP server (`npx @cloudbase/cloudbase-mcp@latest`) or the Open Plugin Spec plugin — inferred from the 'any MCP-capable tool' instructions, not officially documented for DSH
MCP-driven backend operations
natural-language requests in any MCP-capable AI tool→CloudBase login, data query and changes, cloud function and hosting management, log reading — all from the conversation
Writes to live CloudBase resources (data, functions, hosting) in your environmentRequires CloudBase login credentialsPlugin + Skills one-shot setup
`npx plugins add TencentCloudBase/cloudbase-plugin` in an Open Plugin Spec tool→MCP Server, Agent Skills and Hooks installed together — scenario skills for Web / Mini Program / database / auth / functions
Installs MCP Server, Skills and Hooks assets into your AI tool's configurationDatabase, compute, auth and release coverage
backend requirements described in chat (models, functions, login methods, releases)→PostgreSQL / document DB models and CRUD with security rules; cloud functions / Cloud Run authoring, deploy, invoke, debug; auth and object storage; static hosting / Mini Program publish with logs and redeploy
Deploys code and publishes apps to your CloudBase environmentCloudBase usage beyond free quotas is billed
04Who is it for? When not to use it?
Good for
- Developers using MCP-capable AI coding tools who want Tencent CloudBase backend work done in chat instead of the cloud console
- Operators and indie builders maintaining CloudBase apps who want data CRUD, deploys and log reads handled by the agent
Not for
- Requires your own CloudBase environment, and sensitive actions proposed by the AI must be confirmed manually — the Toolkit provides capability and path, not judgment.
- The README does not document DeepSeek Harness setup; using it from DSH follows the generic 'Other MCP-capable IDEs' MCP-config path and is not officially verified.
05Compatibility, maintenance and safety notes
- Requires your own CloudBase environment, and sensitive actions proposed by the AI must be confirmed manually — the Toolkit provides capability and path, not judgment.
- The Toolkit itself is MIT open source, but CloudBase has free quotas only — usage beyond that is billed by Tencent Cloud.
- The README does not document DeepSeek Harness setup; using it from DSH follows the generic 'Other MCP-capable IDEs' MCP-config path and is not officially verified.
MIT · actively maintained (latest release v2.31.0, 2026-08-20)
06Frequently asked questions
How do I connect CloudBase AI Toolkit to DeepSeek Harness?
The README does not document DeepSeek Harness specifically. Since the toolkit works with any MCP-capable tool, in DSH you can configure an MCP server with command `npx` and args `@cloudbase/cloudbase-mcp@latest`, or install the Open Plugin Spec plugin `TencentCloudBase/cloudbase-plugin` if DSH supports it. This path is inferred from the generic MCP instructions, not officially verified.
Is the DeepSeek Harness integration native or MCP?
MCP / Plugin — there is no native DeepSeek Harness runtime. The Plugin installs MCP Server, Agent Skills and Hooks together; MCP-only setup is a JSON config block in your tool.
What do I need before installing?
Node.js v18.15.0+, a CloudBase environment, and an AI tool that supports Plugin / Skills / MCP.
Where do my code and data go?
Deploys target your own CloudBase environment. In local mode the MCP server runs on your machine and code need not leave until you deploy; cloud traffic uses HTTPS.
Does it cost anything?
The Toolkit (including MCP) is open source under MIT. CloudBase has free quotas; usage beyond that is billed.
07Related DSH workflows
ruflo
by ruvnet
🌊 The original agent meta-harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
reactive-resume
by amruthpillai
A one-of-a-kind resume builder that keeps your privacy in mind. Completely secure, customizable, portable, open-source and free forever. Try it out today!
everos
by evermind-ai
One portable memory layer for every AI agent: local-first, Markdown-native, user-owned, and self-evolving across apps, tools, and workflows.
yao
by yaoapp
✨ All your agents and workspaces in one place, on every device you own. Track tasks on a board, accessible from desktop, mobile, browser, or API. Self-hosted.
08Data and sources
| Other MCP-capable IDEs | MCP config only (below) |
Yes. Any tool that can configure an MCP Server or install the matching Plugin / Skills works—including Claude Code, Gemi…
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-08-21. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.