- LIVE
- STAGED
- ERROR
- HUMAN
Short answer
How can I deploy services using my AI?
You provide your AI with a YAML template or a ZIP package, and it manages the deployment process through Zeabur. It handles everything from preparing the files to launching the service on the cloud.
Deployment outcomes
Where your services land.
What happens after you give your AI the deployment instructions.
LIVE
Active cloud services
The AI finishes the deployment process. You get a running service accessible via the web.
STAGED
Ready for review
The AI uploads your files to a staging area. You can check the contents before the final launch.
ERROR
Deployment failures
If the configuration is broken, the AI identifies the issue. You'll need to fix the code or template manually.
HUMAN
Manual intervention
The AI handles the heavy lifting, but you stay in control of the final approval for production pushes.
The workflow
What your AI does when the deployment request arrives.
The AI moves through a specific sequence to turn your files into running infrastructure.
Prepare the files
The AI organizes your code or configuration into a format Zeabur can process.
prepare_deploymentUpload to staging
Your files are sent to a secure upload stage to ensure they are ready for the cloud.
create_upload_stageExecute deployment
The AI triggers the actual launch using your provided configuration.
deploy_templateVerify status
The AI monitors the deployment to ensure the service reaches a running state.
deploy_template
Try it
Copy these to start.
Use these prompts to kick off a deployment.
Starting points
Swap out the bracketed text with your actual file names or configuration details.
Zeabur Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_C259WqFoZVL0U3fl3UAYmeTpTKhOp70hGTqhzXJ5/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Zeabur capabilities are ready to use.
{
"mcpServers": {
"zeabur-paas-deployment-mcp": {
"url": "https://edge.vinkius.com/vk_preview_C259WqFoZVL0U3fl3UAYmeTpTKhOp70hGTqhzXJ5/mcp"
}
}
}Deploy this service using the YAML template in my current directory.
Create a deployment stage for this ZIP package containing my web app.
Use the provided configuration to launch a new service on Zeabur.
Prepare a deployment for the code I just uploaded.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Zeabur once, then ask.
Just link your account once. Your credentials stay encrypted, and you can start deploying immediately through your AI client.
Connect Zeabur to your AIFAQ
How the deployment behaves
- 01
Can my AI deploy services from a GitHub repo?
No. The AI currently deploys using YAML templates or ZIP packages provided in the chat context.
- 02
Does the AI handle billing for the cloud services?
No. The AI manages the deployment process, but all billing and subscription management happens directly through Zeabur.
- 03
Can I ask the AI to delete a service?
No. The available tools only cover preparing and executing deployments, not managing or deleting existing services.
- 04
What happens if my YAML file has a syntax error?
The deployment will fail. The AI will report the error, and you will have to fix the file before trying again.
- 05
Can the AI deploy to AWS or Google Cloud instead?
No. This specific connector is built to work with Zeabur's deployment infrastructure.
More questions about Zeabur? The Connector page answers them. See everything the Zeabur Connector can do
Connect Zeabur to Claude, Cursor, ChatGPT & more
