PlugLayer MCP Server
Enables deploying and managing infrastructure via natural language, including project/domain management, compute nodes, image deployment, and CI/CD integration.
README
PlugLayer MCP Server
Deploy and manage your infrastructure through natural language with any MCP-compatible AI assistant.
Installation
Option 1: uvx (recommended — no install needed)
PLUGLAYER_API_KEY=your-pluglayer-api-token uvx pluglayer-mcp
This local command mode uses the MCP stdio transport by default, which is the right mode for Cursor, Claude Code, and other editor-launched command servers.
The pluglayer-mcp command now always uses stdio so editor clients cannot accidentally switch it into HTTP mode.
Option 2: pip
pip install pluglayer-mcp
PLUGLAYER_API_KEY=your-pluglayer-api-token pluglayer-mcp
Configuration
Claude Desktop
Add to ~/.config/Claude/claude_desktop_config.json:
{
"mcpServers": {
"pluglayer": {
"command": "uvx",
"type": "stdio",
"args": ["pluglayer-mcp"],
"env": {
"PLUGLAYER_API_KEY": "your-pluglayer-api-token"
}
}
}
}
Cursor
Add to ~/.cursor/mcp.json:
{
"pluglayer": {
"command": "uvx",
"type": "stdio",
"args": ["pluglayer-mcp"],
"env": {
"PLUGLAYER_API_KEY": "your-pluglayer-api-token"
}
}
}
Remote HTTP (hosted)
The remote MCP server runs at mcp.pluglayer.com. Pass your token as:
Authorization: Bearer your-pluglayer-api-token
If you intentionally want to run the package itself as an HTTP MCP server, use:
pluglayer-mcp-http
Release Checklist
Before publishing a new pluglayer-mcp build:
- Confirm local command mode still uses
stdioby default. - Confirm
PLUGLAYER_API_URLoverride works when pointed at a dev API. - Confirm package version will be unique for the publish run.
- Publish from the public repo
mainbranch after reviewing thedev -> mainPR.
After publishing:
- Restart Cursor, Claude Code, or the MCP client you are testing.
- If the editor still behaves like an older MCP build, remove and re-add the MCP server entry, then restart the editor.
- Re-test with a simple command such as:
get_current_userlist_projectsget_compute_summary
Cursor Notes
- For
command-based MCP setup, useuvx pluglayer-mcp. - Do not force HTTP transport for local editor usage.
pluglayer-mcpalways usesstdio, which is the correct transport for Cursor-launched command servers.- Only use
pluglayer-mcp-httpwhen you intentionally want to run the package itself as an HTTP MCP server.
Available Tools
The MCP calls the PlugLayer FastAPI backend instead of re-implementing backend business logic. Auth, roles, ownership, compute guards, k3s orchestration, and admin checks remain in the backend. MCP and editor plugins should authenticate with a PlugLayer API token created in the PlugLayer Settings page, not the browser/session auth token.
Managed registries are configured by PlugLayer admins in the platform UI/API. When deploy_image uses mirroring, the backend picks a registry the current user is allowed to use and keeps Kubernetes pull secrets in sync automatically.
| Tool | Description |
|---|---|
get_current_user |
Show the Authentik-backed user and roles |
list_projects |
List authenticated user's projects |
create_project |
Create a new project namespace |
get_project |
Get project details |
get_compute_summary |
Show account-level personal + shared compute capacity |
list_nodes |
List accessible compute nodes |
add_node_ssh |
Add a personal SSH node usable by all of the user's projects |
list_registries |
List the registries currently available to the user |
deploy_image |
Mirror a Docker image into PlugLayer's managed Docker Hub namespace, then deploy it after backend compute checks |
deploy_compose |
Deploy from docker-compose.yml after backend compute checks |
list_deployments |
List running apps/deployments |
get_deployment_status |
Check app status and URL |
get_logs |
Get app logs |
redeploy |
Redeploy an app |
rollback |
Roll back to previous version |
delete_deployment |
Delete an app |
list_project_domains |
List custom domains for a project |
add_custom_domain |
Add a single or wildcard custom domain and return DNS records |
verify_custom_domain |
Verify TXT/CNAME DNS and activate if attached |
attach_custom_domain |
Attach a verified custom domain to an app |
detach_custom_domain |
Detach a domain while keeping verification |
remove_custom_domain |
Remove a domain and its route |
get_task_status |
Poll async operation progress |
admin_get_overview |
Admin-only platform summary |
admin_set_compute_defaults |
Admin-only default compute quota metadata |
admin_set_node_shared |
Admin-only mark node shared/private |
admin_add_shared_ssh_node |
Admin-only add shared PlugLayer SSH compute |
generate_github_actions |
Get CI/CD pipeline YAML |
get_cluster_health |
Check cluster status |
Example Conversations
Deploy your first app:
"I have a FastAPI app at
ghcr.io/myorg/api:latestthat runs on port 8000. Deploy it to myproductionproject."
Convert docker-compose:
"Here's my docker-compose.yml: [paste]. Deploy this to PlugLayer."
Add a node:
"Add my server at 192.168.1.100 as personal compute. Here's my SSH key: [paste]"
CI/CD setup:
"Generate a GitHub Actions workflow for my
apideployment so it auto-deploys on push to main."
Add a custom domain:
"Add
api.example.comto my production project, show me the DNS records, then verify it and attach it to my API app."
Getting Your API Key
- Go to PlugLayer Settings
- Create a PlugLayer API token
- Copy it once and store it safely
- Use it as
PLUGLAYER_API_KEYfor MCP, editor plugins, and CI/CD webhook deploys
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。