Token导航 LogoToken导航TokenDH.com
研究检索external-servicegithub未标认证来源可访问许可证需确认审计提醒

mcp-statusMCP 状态

Agent Skill

mcp-status 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

198

周安装

8

GitHub Stars

公开资料未说明

下载量

62
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:mcp-status(MCP 状态)
来源仓库:https://github.com/fancy-bread/sdlc-workflow-skills
仓库路径:skills/mcp-status
安装命令:
npx skills add https://github.com/fancy-bread/sdlc-workflow-skills --skill mcp-status
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/fancy-bread/sdlc-workflow-skills --skill mcp-status

简介

mcp-status 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中快速定位候选结果。

  • 适用于需要跟踪软件交付流程或验证发布状态的场景。
  • 通过 CI/CD 集成和状态聚合来提供实时视图,具体用法需结合原始 README 确认。
  • 安装前建议确认构建系统权限和维护状态,注意可能涉及 webhook 和 API 调用。
  • 使用时应区分不同环境的状态,避免混淆开发、测试和生产数据。

SKILL.md

MCP Status

Overview

Check the authentication status of all configured Model Context Protocol (MCP) servers.

Definitions

  • MCP server: A configured Model Context Protocol server in Cursor (e.g. github, atlassian, ado, asdlc).
  • User-level MCP configuration: The mcpServers section in ~/.cursor/mcp.json (macOS/Linux) or %USERPROFILE%\.cursor\mcp.json (Windows), or configured via Cursor Settings → Features → Model Context Protocol.
  • Project-level MCP configuration: Optional mcpServers in .cursor/mcp.json at the workspace root. When present, Cursor may merge or override with user-level config; discover from this file so status reflects project-configured servers. If the file is missing, skip (not an error).
  • Extension-exposed MCP: An MCP server provided by a VS Code/Cursor extension (e.g. Agent Context Explorer / extension-ace). Tools may use prefixes like mcp_<vendor>_*. When the agent has access to such tools, include them in status and tag as source (extension).

Prerequisites

  • None required. Run anytime to check status. If no MCP servers are configured, the command reports that.
  • MCP Tool Usage Standards: MCP tool usage should follow best practices (check schema files, validate parameters, handle errors gracefully). These standards are documented in AGENTS.md §3 Operational Boundaries if AGENTS.md exists, but apply universally regardless.

Purpose

MCP servers can disconnect or lose authentication after periods of inactivity. Use this command to verify all integrations are ready before starting work.

Steps

  1. Discover configured MCP servers (user, project, extension)

- User-level: Read ~/.cursor/mcp.json (macOS/Linux) or %USERPROFILE%\.cursor\mcp.json (Windows). Extract mcpServers keys. Tag each server as source (user). If the file is not accessible, try common server names: github, atlassian, ado, asdlc, user-github, user-atlassian, user-ado, user-asdlc. - Project-level: If workspace root is available, read .cursor/mcp.json at workspace root. If the file exists, extract mcpServers keys and tag each as (project). If a server name already appeared from user config, record both sources (e.g. "user, project") or report once with combined source. If the file is missing, skip (not an error). - Extension-exposed: When the agent has access to MCP tools from extensions (e.g. Agent Context Explorer; tool names use a prefix pattern such as mcp_<vendor>_<suffix>), treat each distinct extension server as one entry. Call one read-only tool per known extension server (e.g. extension-ace: list rules or list commands) and tag as (extension). If no extension tools are available, skip (not an error). - Note: Server names in config may differ from tool prefixes (e.g. config has github but tools use mcp_github_*).

  1. Test each server connection

- For each discovered server (from any source), attempt to call one lightweight read-only tool to verify connectivity and authentication. Record the server's source (user / project / extension) with the result. - Use common tool patterns for known server types: - github / user-github → Try list_commits (may require owner/repo args) or list_branches - atlassian / user-atlassian → Try getAccessibleAtlassianResources or atlassianUserInfo - ado / user-ado → Try core_list_projects - asdlc / user-asdlc → Try list_articles - For unknown server types, try common tool names like list_*, get_*, or *_info with minimal or empty args. - Record success or failure for each server. Handle "server not found" vs "authentication error" vs "tool not found" differently.

  1. Report status

- Display results in a clear, formatted list. Group or label by source (User config, Project config, Extensions) so users see where each server comes from. - Show server name, authentication status, and source (e.g. "(user)", "(project)", "(extension)"). - For disconnected servers, provide reconnection instructions (Cursor Settings → Features → Model Context Protocol).

Tools

Filesystem

  • Read user MCP configuration: ~/.cursor/mcp.json (macOS/Linux) or %USERPROFILE%\.cursor\mcp.json (Windows). Parse JSON to extract mcpServers keys.
  • Read project MCP configuration: .cursor/mcp.json at workspace root (if present). Parse JSON to extract mcpServers keys. If file is missing, skip.

MCP (per discovered server)

  • github / user-github → Try list_commits, list_branches, or other read-only tools
  • atlassian / user-atlassian → Try getAccessibleAtlassianResources, atlassianUserInfo
  • ado / user-ado → Try core_list_projects
  • asdlc / user-asdlc → Try list_articles
  • Extension-exposed (e.g. extension-ace): Try a read-only tool such as list rules or list commands (no args or minimal args). Tag result as (extension).
  • Other servers: Try common read-only tool patterns (list_*, get_*, *_info) with minimal or empty args
  • Note: Tool names may be prefixed with mcp_<server>_ or mcp_user-<server>_ depending on configuration. Record source (user / project / extension) with each result.

Expected Output

All Connected (with sources)

🔌 MCP Server Status

User config:
  ✅ atlassian - Connected (user)
  ✅ github - Connected (user)

Project config:
  (none)

Extensions:
  ✅ extension-ace - Connected (extension)

All systems operational!

(When only user-level config exists and no extensions, output may show no "Project config" or "Extensions" sections, and servers with "(user)" or no source label for backward compatibility.)

Some Disconnected

🔌 MCP Server Status

User config:
  ❌ atlassian - Needs authentication (user)
  ✅ github - Connected (user)

⚠️ Action Required:
1. Open Cursor Settings (Cmd+, or Ctrl+,)
2. Navigate to: Tools & MCP
3. Click "Connect" next to: atlassian
4. Run /mcp-status again to verify

When to Use

  • Start of day - Verify connections before beginning work
  • After inactivity - MCP servers may disconnect after timeout
  • Before critical commands - Ensure integrations are ready for commands like /start-task, /create-task, etc.
  • Troubleshooting - When other commands fail with authentication errors

Error Handling

If unable to discover MCP servers:

  • If user config file is not accessible, try common server names as fallback
  • If project .cursor/mcp.json is missing, skip (not an error)
  • If no extension tools are available, skip extension section (not an error)
  • If no servers respond, report that no MCP servers are configured or accessible
  • Provide link to MCP setup documentation (e.g., docs/reference/mcp-setup.md if present, or general MCP setup instructions)

If a server test fails:

  • Server not found: Server name doesn't exist in MCP configuration
  • Authentication error: Server exists but needs reconnection/authentication
  • Tool not found: Server exists but the tested tool isn't available (try a different tool)
  • Network/connection error: Server unreachable or connection failed
  • Provide specific guidance for each failure type, especially authentication errors which require user action

Notes

  • This command performs read-only operations only
  • No data is modified or created
  • Safe to run at any time
  • Does not require any parameters or arguments

Guidance

Role

Act as a developer checking that MCP integrations are ready before running commands that depend on them.

Instruction

Read the MCP configuration file (~/.cursor/mcp.json or Windows equivalent) to discover configured servers. For each server, attempt to call a lightweight read-only MCP tool to verify connectivity and authentication. Report connected / disconnected status; for disconnected servers, provide reconnection steps (Cursor Settings → Features → Model Context Protocol).

Context

  • MCP servers can disconnect or lose auth after inactivity. Use at start of day, after inactivity, or before critical commands.
  • Discover servers from three sources: (1) user-level mcp.json, (2) project-level .cursor/mcp.json at workspace root when present, (3) extension-exposed MCPs when the agent has access to their tools. Report each server with its source (user / project / extension).
  • If user config file is not accessible, try common server names and test connectivity. Missing project or extension config is not an error.
  • ASDLC patterns: Context Gates
  • ASDLC pillars: Quality Control (pre-flight validation for other commands)

Examples

ASDLC: Context Gates — MCP checks act as an input gate before running commands that depend on them.

Constraints

Rules (Must Follow):

  1. Operational Standards Compliance: This command follows operational standards (documented in AGENTS.md if present, but apply universally):

- MCP Tool Usage: Check schema files, validate parameters, handle errors gracefully - AGENTS.md Optional: Commands work without AGENTS.md. Standards apply regardless of whether AGENTS.md exists. - See AGENTS.md §3 Operational Boundaries (if present) for detailed standards

  1. Read-only Operations: This command performs read-only operations only; no data is modified or created.
  2. Error Handling: If a server test fails, distinguish authentication errors (needs reconnect) from other errors and provide specific guidance.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

37.64%
按下载量换算23

Claude

32.66%
按下载量换算20

Cursor

16.73%
按下载量换算10

Gemini CLI

9.3%
按下载量换算6

安全审计

Gen Agent Trust Hub

可疑

Socket

通过

Snyk

可疑

权限和风险

external-service

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills