Token导航 LogoToken导航TokenDH.com
研究检索敏感数据clawhub未标认证来源可访问clear审计通过

mcp-server-setup-kitMCP server 设置 KIT

Agent Skill

mcp-server-setup-kit 用于查找、检索和筛选相关信息,适合在 OpenClaw 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

8,126

周安装

342

GitHub Stars

公开资料未说明

下载量

2,845
OpenClaw

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

MIT-0

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-server-setup-kit(MCP server 设置 KIT)
来源仓库:https://github.com/flynndavid/mcp-server-setup-kit
安装命令:
openclaw skills install mcp-server-setup-kit
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 OpenClaw 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。

ClawHubOpenClaw
openclaw skills install mcp-server-setup-kit

简介

指导您使用经过测试的 MCP 服务器设置协议以及配置模板和测试,逐步将 Claude 连接到 Notion、Linear、Slack 和 GitHub。

SKILL.md

name
mcp-server-setup-kit
version
1.0.0
price
19
bundle
ai-setup-productivity-pack
bundle_price
79
last_validated
2026-03-07

MCP Server Setup Kit

Framework: The 5-Minute Connect Protocol *Worth $200/hr consultant time. Yours for $19.*


What This Skill Does

Guides you through connecting Claude to Notion, Linear, Slack, and GitHub using the Model Context Protocol (MCP) — in one focused workflow. No trial-and-error. No missing steps. Just a working integration you can test in under 5 minutes per tool.

Problem it solves: MCP setup friction is the #1 reason teams abandon agent workflows in 2026. The docs exist but the path is scattered. This skill gives you the straight line.


The 5-Minute Connect Protocol

A structured checklist that takes any MCP server from "never heard of it" to "Claude is using it" in 5 minutes or less.

Phase 1 — Qualify (30 seconds)

Answer these before touching any config:

QuestionYes → ContinueNo → Fix First
Do you have a Claude Desktop or OpenClaw installation?Install first
Do you have an API key / OAuth token for the target tool?Generate it now
Do you know where your claude_desktop_config.json lives?Find it (see below)
Is Node.js 18+ or Python 3.10+ installed?Install via homebrew/nvm

Config file locations:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
  • OpenClaw: ~/.openclaw/openclaw.json (mcpServers block)

Phase 2 — Install (2 minutes)

Universal install pattern:

# For NPX-based MCP servers (most common)
npx -y @modelcontextprotocol/server-{toolname}

# For Python-based MCP servers
pip install mcp-server-{toolname}
uvx mcp-server-{toolname}

Verify the binary works before touching config:

npx -y @modelcontextprotocol/server-github --help
# Should print usage/options — if it errors, fix here before config

Phase 3 — Configure (1 minute)

Add the server block to your config. Universal template:

{
  "mcpServers": {
    "{tool-name}": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-{tool-name}"],
      "env": {
        "{TOOL}_API_KEY": "your-key-here"
      }
    }
  }
}

Phase 4 — Test (90 seconds)

Use these verification prompts immediately after restart:

"List my available MCP tools"
"What can you do with [tool-name]?"
"[Tool-specific test prompt from templates below]"

If Claude doesn't see the tool: restart Claude Desktop / OpenClaw gateway completely (not just refresh).


Phase 5 — Validate (30 seconds)

✅ Claude lists the tool when asked ✅ Tool-specific test prompt returns real data ✅ Write operation (if applicable) succeeds ✅ No auth errors in logs

Log check: ~/Library/Logs/Claude/mcp*.log (macOS)


5 Pre-Built Integration Templates

Template 1: GitHub MCP

Use case: Let Claude read repos, issues, PRs, and push code.

{
  "mcpServers": {
    "github": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-github"],
      "env": {
        "GITHUB_PERSONAL_ACCESS_TOKEN": "ghp_your_token_here"
      }
    }
  }
}

Token scopes needed: repo, read:user, read:org Generate at: github.com → Settings → Developer Settings → Personal Access Tokens

5-Minute Connect test prompts:

  1. "List my open GitHub issues across all repos"
  2. "What PRs are waiting for my review?"
  3. "Show me the README for [your-repo]"

Template 2: Notion MCP

Use case: Let Claude read/write Notion pages and databases.

{
  "mcpServers": {
    "notion": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-notion"],
      "env": {
        "NOTION_API_TOKEN": "secret_your_token_here"
      }
    }
  }
}

Token setup: notion.com → Settings → Connections → Develop integrations → New integration Critical step: Share target pages with your integration (Notion doesn't auto-grant access)

5-Minute Connect test prompts:

  1. "List my Notion pages"
  2. "Search Notion for [topic]"
  3. "Create a new page titled 'MCP Test' in [workspace]"

Template 3: Slack MCP

Use case: Let Claude read channels, send messages, search history.

{
  "mcpServers": {
    "slack": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-slack"],
      "env": {
        "SLACK_BOT_TOKEN": "xoxb-your-token",
        "SLACK_TEAM_ID": "T0XXXXXXX"
      }
    }
  }
}

Token setup: api.slack.com → Create App → OAuth & Permissions Scopes needed: channels:read, chat:write, channels:history, users:read

5-Minute Connect test prompts:

  1. "List my Slack channels"
  2. "What was discussed in #general today?"
  3. "Send 'MCP connected!' to #general"

Template 4: Linear MCP

Use case: Let Claude read/create/update Linear issues and projects.

{
  "mcpServers": {
    "linear": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-linear"],
      "env": {
        "LINEAR_API_KEY": "lin_api_your_key_here"
      }
    }
  }
}

Key setup: linear.app → Settings → API → Personal API Keys

5-Minute Connect test prompts:

  1. "What issues are assigned to me in Linear?"
  2. "Show my current sprint"
  3. "Create a new Linear issue: 'Test MCP connection' in [Team]"

Template 5: Multi-Tool Stack (All 4 at Once)

For teams who want the full setup in one shot:

{
  "mcpServers": {
    "github": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-github"],
      "env": { "GITHUB_PERSONAL_ACCESS_TOKEN": "ghp_..." }
    },
    "notion": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-notion"],
      "env": { "NOTION_API_TOKEN": "secret_..." }
    },
    "slack": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-slack"],
      "env": {
        "SLACK_BOT_TOKEN": "xoxb-...",
        "SLACK_TEAM_ID": "T0..."
      }
    },
    "linear": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-linear"],
      "env": { "LINEAR_API_KEY": "lin_api_..." }
    }
  }
}

Cross-tool test prompt: "Using my connected tools: summarize my open Linear issues, find related Notion docs, check GitHub for open PRs, and post a summary to #standup in Slack"


Troubleshooting Decision Tree

Claude doesn't see my MCP tool
├── Did you restart Claude completely? (not just refresh)
│   └── No → Restart fully → Try again
├── Is the config JSON valid?
│   └── Check: jsonlint.com → paste your config
├── Does the server binary work standalone?
│   └── Run: npx -y @modelcontextprotocol/server-{name} --help
│       ├── Error → npm/node version issue → upgrade node
│       └── Works → config path or key issue
├── Auth error in logs?
│   └── Yes → Regenerate API key → check scopes
└── Tool shows but returns empty data?
    └── Notion: did you share pages with integration?
    └── Slack: check bot scopes
    └── GitHub: check token permissions

Scoring Rubric: Are You Connected?

Score your setup after completing Phase 5:

CheckPoints
Tool appears in Claude's tool list20
Read test prompt returns real data30
Write test prompt succeeds30
No errors in MCP logs10
Cross-tool prompt works (multi-stack)10

80-100: Fully connected. Ship it. 60-79: Partial connection. Check scopes and restart. Below 60: Go back to Phase 2. Something in install/config is broken.


Example Session

User prompt:

"Use this skill to connect me to GitHub"

Agent response:

  1. Runs Phase 1 checklist aloud (asks for token if missing)
  2. Provides Template 1 config block (pre-filled with user's token)
  3. Instructs restart
  4. Runs test prompts and confirms output
  5. Scores the setup using the rubric
  6. Suggests next integration (Notion → Slack → Linear)

Bundle Note

This skill is part of the AI Setup & Productivity Pack ($79 bundle):

  • MCP Server Setup Kit ($19) — *you are here*
  • Agentic Loop Designer ($29)
  • AI OS Blueprint ($39)
  • Context Budget Optimizer ($19)
  • Non-Technical Agent Quickstart ($9)

Save $36 with the full bundle. Built by @Remy_Claw.

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

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

补充不同宿主或平台的使用分布数据

能力 5

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

OpenClaw

83.36%
按下载量换算2,372

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

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。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