阿里云MCP服务器 是 AI Skill Hub 本期精选MCP工具之一。综合评分 8.0 分,整体质量较高。我们强烈推荐将其纳入你的 AI 工具库,帮助提升工作效率。
阿里云MCP服务器 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
阿里云MCP服务器 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
# 方式一:通过 Claude Code CLI 一键安装
claude skill install https://github.com/aliyun/alibabacloud-devops-mcp-server
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"---mcp---": {
"command": "npx",
"args": ["-y", "alibabacloud-devops-mcp-server"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
# 安装后在 Claude 对话中直接使用 # 示例: 用户: 请帮我用 阿里云MCP服务器 执行以下任务... Claude: [自动调用 阿里云MCP服务器 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入:"列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"___mcp___": {
"command": "npx",
"args": ["-y", "alibabacloud-devops-mcp-server"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
AlibabaCloud Devops MCP Server provides AI assistants with the ability to interact with the Yunxiao platform, enabling them to read work item contents in projects, automatically write code after understanding requirements, and submit code merge requests. Enterprise development teams can use it to assist with code reviews, optimize task management, reduce repetitive operations, and thus focus on more important innovation and product delivery.
alibabacloud-devops-mcp-server provides the following capabilities for AI assistants:
organization-management: Organization management tools (organization list, organization details, department information, organizational roles, member information, etc.)code-management: Code repository management tools (repository management, branch management, merge request management, file tree, etc.)project-management: Project management tools (project management, work item management, work item fields, work item comments, time tracking, etc.)pipeline-management: Pipeline management tools (pipeline list, pipeline configuration, resource management, tag management, deployment management, etc.)application-delivery: Application delivery tools (deployment order management, application management, application tags, variable group management, etc.)packages-management: Artifact repository management tools (artifact repositories, artifact lists, etc.)test-management: Test management tools (test case management, test case directories, test plans, test results, etc.)
Yunxiao provides an officially hosted MCP endpoint — no local install or Docker needed. Just point your client at it and authenticate with your Yunxiao token.
https://openapi-rdc.aliyuncs.com/ai/mcp/ai/mcp, e.g. https://<your-org>.devops.aliyuncs.com/ai/mcp — the same domain you use to access Yunxiao in your region.Authorization: Bearer <YOUR_TOKEN> (or header X-Yunxiao-Token: <YOUR_TOKEN>)Client config — Central (clients with native remote support, e.g. Cursor):
{
"mcpServers": {
"yunxiao": {
"url": "https://openapi-rdc.aliyuncs.com/ai/mcp",
"headers": { "Authorization": "Bearer <YOUR_TOKEN>" }
}
}
}
Client config — Region edition (replace the URL with your own region domain):
{
"mcpServers": {
"yunxiao": {
"url": "https://<your-org>.devops.aliyuncs.com/ai/mcp",
"headers": { "Authorization": "Bearer <YOUR_TOKEN>" }
}
}
}
Tip: cut context usage by requesting only the toolsets you need —?toolsets=code-management,project-managementor headerX-Devops-Toolsets. See the full guide: docs/hosted-mcp-guide.zh-CN.md.
---
cp .env.example .env # set YUNXIAO_ACCESS_TOKEN
docker compose up -d
Client URL: http://localhost:3000/sse (SSE) or http://localhost:3000/mcp (Streamable HTTP)
Stdio mode is the simplest self-hosting way, suitable for most MCP clients (like Cursor, Claude Desktop, iFlow, etc.). No Docker installation required, just run via npx.
Add the following configuration to your MCP client configuration file:
{
"mcpServers": {
"yunxiao": {
"command": "npx",
"args": [
"-y",
"alibabacloud-devops-mcp-server"
],
"env": {
"YUNXIAO_ACCESS_TOKEN": "<YOUR_TOKEN>"
}
}
}
}
Note: - Replace<YOUR_TOKEN>with your Yunxiao access token - The-yflag automatically confirms installation without manual confirmation - By default the tool calls the Yunxiao Central Station OpenAPI (https://openapi-rdc.aliyuncs.com); no extra configuration is required - If you are on a Region edition (organization-specific domain), set an additional environment variableYUNXIAO_API_BASE_URLto your Yunxiao instance URL, e.g.https://your-org.devops.aliyuncs.com. See the Configuring Region Edition section above for details - This method uses stdio mode, communicating with the MCP client via standard input/output
---
Stdio:
{
"mcpServers": {
"yunxiao": {
"command": "docker",
"args": ["run", "-i", "--rm", "-e", "YUNXIAO_ACCESS_TOKEN", "build-steps-public-registry.cn-beijing.cr.aliyuncs.com/build-steps/alibabacloud-devops-mcp-server:latest"],
"env": { "YUNXIAO_ACCESS_TOKEN": "<YOUR_TOKEN>" }
}
}
}
SSE: http://localhost:3000/sse
Streamable HTTP: http://localhost:3000/mcp
Pass credentials via query parameter or header:
http://localhost:3000/mcp?yunxiao_access_token=YOUR_TOKEN_HERE Or request header: x-yunxiao-token: YOUR_TOKEN_HERE
Region / instance OpenAPI base:
http://localhost:3000/mcp?yunxiao_access_token=TOKEN&yunxiao_api_base_url=https%3A%2F%2Fyour-org.devops.aliyuncs.com Or request header: x-yunxiao-api-base-url: https://your-org.devops.aliyuncs.com
get_pipeline: Get pipeline detailslist_pipelines: Get pipeline listsmart_list_pipelines: Smart pipeline search with natural language time referencesgenerate_pipeline_yaml: [Pipeline Management] Generate pipeline YAML configurationcreate_pipeline_from_description: Create a pipeline from natural language descriptionupdate_pipeline: [Pipeline Management] Update pipeline YAML contentcreate_pipeline_run: Create a pipeline run instanceget_latest_pipeline_run: Get the latest pipeline run instanceget_pipeline_run: Get pipeline run detailslist_pipeline_runs: Get pipeline run listlist_pipeline_jobs_by_category: Get pipeline execution tasks by categorylist_pipeline_job_historys: Get the execution history of a pipeline taskexecute_pipeline_job_run: Manually run a pipeline taskget_pipeline_job_run_log: Get the execution logs of a pipeline joblist_service_connections: List service connections in organizationcreate_pipeline_from_description: Automatically generates YAML configuration and creates pipelineupdate_pipeline: Update an existing pipeline in Yunxiao by pipelineId. Use this to update pipeline YAML, stages, jobs, etc.create_resource_member: Create a resource memberdelete_resource_member: Delete a resource memberlist_resource_members: Get a list of resource membersupdate_resource_member: Update a resource memberupdate_resource_owner: Transfer resource ownercreate_tag: Create a tagcreate_tag_group: Create a tag grouplist_tag_groups: Get a list of tag groupsdelete_tag_group: Delete a tag groupupdate_tag_group: Update a tag groupget_tag_group: Get a tag groupdelete_tag: Delete a tagupdate_tag: Update a tagstop_vm_deploy_order: Stop VM deploy orderskip_vm_deploy_machine: Skip VM deploy machineretry_vm_deploy_machine: Retry VM deploy machineresume_vm_deploy_order: Resume VM deploy orderget_vm_deploy_order: Get VM deploy order detailsget_vm_deploy_machine_log: Get VM deploy machine loglist_package_repositories: Get package repositories details listlist_artifacts: Get artifacts details listget_artifact: Get single artifact detailsalibabacloud-devops-mcp-server 是专为 AI 助手设计的 MCP Server,旨在通过 Model Context Protocol (MCP) 将 AI 能力与 AlibabaCloud Devops(云效)平台深度连接。它允许 AI 助手直接读取项目中的 Work Item 内容、理解需求并自动编写代码,甚至能够自主提交 Merge Request。通过该工具,企业开发团队可以实现智能化的代码评审、任务管理优化以及研发流程自动化,显著提升开发效率。
本项目为 AI 助手提供了丰富的管理能力:首先是 organization-management(组织管理),支持查询组织列表、部门信息、角色及成员详情;其次是 code-management(代码管理),涵盖了代码仓库、分支管理、Merge Request 管理以及文件树查看等功能。通过这些能力,AI 可以无缝介入研发全生命周期。
在使用本项目前,请确保您的开发环境满足以下要求:1. Node.js 版本需大于等于 18.0.0;2. 您需要拥有 AlibabaCloud Devops 的 Personal Access Token(个人访问令牌),并确保该 Token 已授予相应的读(Read)和写(Write)权限,以便 AI 助手能够正常操作云效资源。
您可以根据使用场景选择不同的安装方式:1. 通过 Lingma(通义灵码)内置的 MCP Marketplace 进行安装,搜索 "Yunxiao DevOps" 并点击安装即可;2. 使用 Docker 进行部署,通过 Docker Compose 启动服务,并配置好 .env 文件中的 YUNXIAO_ACCESS_TOKEN;3. 对于开发者,也可以通过 npx 直接运行相关包。
推荐使用 Stdio 模式进行快速启动,这是最简单且通用的方式,适用于 Cursor、Claude Desktop、iFlow 等主流 MCP Client。在 Stdio 模式下,您无需安装 Docker,直接通过 npx 命令即可运行,非常适合快速集成到您的 AI 开发工作流中。
配置过程取决于您选择的运行模式:1. 若使用 NPX (Stdio 模式),请在 MCP Client 的配置文件中添加对应的 JSON 配置,并务必在 env 字段中填入您的 YUNXIAO_ACCESS_TOKEN;2. 若使用 Docker 运行,请配置相应的环境变量;3. 若通过 SSE 模式接入,请使用提供的 HTTP URL 进行连接。
本项目提供了强大的自动化工具集:Pipeline Management Tools(流水线管理工具)支持获取流水线详情、列表查询,并具备通过自然语言进行智能搜索及根据描述自动生成 Pipeline YAML 的能力;Packages Management Tools(制品管理工具)则支持查询包仓库详情、制品列表及单个制品信息,实现研发资产的智能化管理。
高质量的MCP服务器实现
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ Apache 2.0 — 宽松开源协议,可商用,需保留版权声明和 NOTICE 文件,含专利授权条款。
经综合评估,阿里云MCP服务器 在MCP工具赛道中表现稳健,质量优秀。如果你已有明确的使用需求,可以直接上手体验;如果还在评估阶段,建议对比同类工具后再做决策。
| 原始名称 | alibabacloud-devops-mcp-server |
| 原始描述 | 开源MCP工具:Yunxiao MCP Server provides AI assistants with the ability to interact with the 。⭐116 · TypeScript |
| Topics | alibabacloud-devopsdevops-mcptypescript |
| GitHub | https://github.com/aliyun/alibabacloud-devops-mcp-server |
| License | Apache-2.0 |
| 语言 | TypeScript |
收录时间:2026-06-03 · 更新时间:2026-06-05 · License:Apache-2.0 · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端