Argo Workflow MCP Server

Argo Workflow MCP Server

Enables AI agents to manage Argo Workflows through REST API, supporting workflow template and instance operations including creation, submission, monitoring, and deletion with token authentication.

Category
Visit Server

README

Argo Workflow MCP Server

一个 Model Context Protocol (MCP) 服务器,为 AI Agent 提供管理 Argo Workflows 的能力。

功能特性

  • 🚀 通过 Argo Server REST API 管理 Argo Workflows
  • 📝 创建、查询、列出和删除 Workflow Templates
  • ⚡ 提交、监控和管理 Workflow 实例
  • 🔒 支持 Token 认证和 TLS 证书验证
  • 🛠️ 完整的 MCP 协议支持
  • 📦 易于安装和配置

安装

使用 npx(推荐)

无需安装,直接运行:

npx argo-workflow-mcp-server --argo-server https://argo-server.example.com --argo-token your-token

全局安装

npm install -g argo-workflow-mcp-server

使用方法

命令行参数

argo-workflow-mcp-server [options]

选项:
  --argo-server <url>      Argo Server 地址(必需)
  --argo-token <token>     Argo Server 认证 token(可选)
  --argo-insecure          跳过 TLS 证书验证(用于自签名证书或开发环境)
  --namespace <name>       默认命名空间(默认:argo)
  --log-level <level>      日志级别:error, warn, info, debug(默认:info)
  -h, --help               显示帮助信息
  -V, --version            显示版本号

示例

连接到生产环境 Argo Server

argo-workflow-mcp-server \
  --argo-server https://argo-server.example.com \
  --argo-token your-auth-token \
  --namespace argo

连接到本地开发环境(通过 port-forward)

首先设置 port-forward:

kubectl port-forward -n argo svc/argo-server 2746:2746

然后启动 MCP Server:

argo-workflow-mcp-server \
  --argo-server http://localhost:2746 \
  --argo-insecure \
  --namespace argo

连接到使用自签名证书的 Argo Server

如果你的 Argo Server 使用自签名证书,会遇到 "self-signed certificate" 错误。使用 --argo-insecure 参数跳过证书验证:

argo-workflow-mcp-server \
  --argo-server https://argo-server.example.com \
  --argo-token your-auth-token \
  --argo-insecure \
  --namespace argo

注意--argo-insecure 会跳过 TLS 证书验证,仅在以下情况使用:

  • 开发和测试环境
  • 使用自签名证书的内部环境
  • 通过 port-forward 连接本地环境

不推荐在生产环境使用,生产环境应使用有效的 TLS 证书。

AI Agent 配置

Kiro

.kiro/settings/mcp.json 中添加配置:

{
  "mcpServers": {
    "argo-workflow": {
      "command": "npx",
      "args": [
        "-y",
        "argo-workflow-mcp-server",
        "--argo-server",
        "https://argo-server.example.com",
        "--argo-token",
        "your-auth-token",
        "--namespace",
        "argo"
      ]
    }
  }
}

其他 MCP 客户端

参考 MCP 文档 配置您的客户端。

可用工具

MCP Server 提供以下工具:

Workflow Template 管理

  • create_workflow_template - 创建新的 Workflow Template
  • get_workflow_template - 查询指定的 Workflow Template
  • list_workflow_templates - 列出所有 Workflow Templates
  • delete_workflow_template - 删除指定的 Workflow Template

Workflow 实例管理

  • submit_workflow - 从模板提交新的 Workflow
  • get_workflow_status - 查询 Workflow 运行状态
  • list_workflows - 列出所有 Workflows
  • delete_workflow - 删除指定的 Workflow

获取 Argo Server 地址和 Token

获取 Argo Server 地址

如果 Argo Server 已暴露为 LoadBalancer 或 Ingress:

kubectl get svc -n argo argo-server

如果使用 port-forward:

kubectl port-forward -n argo svc/argo-server 2746:2746
# 然后使用 http://localhost:2746

获取认证 Token

创建 ServiceAccount 并获取 token:

# 创建 ServiceAccount
kubectl create sa argo-workflow-mcp -n argo

# 创建 RoleBinding(根据需要调整权限)
kubectl create rolebinding argo-workflow-mcp \
  --clusterrole=argo-admin \
  --serviceaccount=argo:argo-workflow-mcp \
  -n argo

# 获取 token
kubectl create token argo-workflow-mcp -n argo --duration=8760h

开发

构建

npm install
npm run build

测试

npm test

本地测试

npm run build
npm link
argo-workflow-mcp-server --help

故障排查

自签名证书错误

错误信息Failed to connect to Argo Server: self-signed certificate

解决方案:使用 --argo-insecure 参数跳过 TLS 证书验证

argo-workflow-mcp-server \
  --argo-server https://your-argo-server \
  --argo-token your-token \
  --argo-insecure

连接失败

  • 确认 Argo Server 地址正确
  • 检查网络连接
  • 验证 token 是否有效
  • 如果使用 HTTPS,确保证书有效或使用 --argo-insecure

权限错误

  • 确认 ServiceAccount 有足够的权限
  • 检查 RBAC 配置
  • 验证 token 是否过期

更多帮助

查看 故障排查指南

许可证

MIT License - 详见 LICENSE 文件

贡献

欢迎提交 Issue 和 Pull Request!

相关链接

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
E2B

E2B

Using MCP to run code via e2b.

Official
Featured