LeadClaw MCP

LeadClaw MCP

Demo MCP server for booking confirmed sales appointments through Claude. Allows users to request appointments, check status, and get detailed BANT briefings using natural language, with realistic fake data for demonstration.

Category
访问服务器

README

LeadClaw MCP (demo)

Demo MCP server for LeadClaw — a service for booking confirmed sales appointments through Claude. Human freelance SDRs handle the outbound; AI handles matching, quality, billing, and reporting. Customers pay only per BANT-qualified meeting that lands on their calendar.

This server is the showcase implementation used for the demo video and concept pitch. It returns realistic fake data so Claude Desktop calls feel live without any real SDR plumbing behind them yet.


What this demo proves

In Claude Desktop you can say:

Book 3 sales appointments next Tuesday in Dallas with AI startup CEOs. Budget $450.

…and Claude will:

  1. Call book_appointments on this server.
  2. Get back a request_id and an ETA.
  3. Poll check_status (immediately or later) and see appointments populate.
  4. Pull full BANT briefings with get_appointment_details.

Nothing real happens on the SDR side — but the customer experience is identical to what the real product will feel like.


Tools exposed

Tool Purpose
book_appointments Submit a request for N confirmed appointments. Returns a request_id.
check_status Poll a request; returns confirmed appointments so far and current phase.
get_appointment_details Full briefing for one appointment: contact, company, BANT, SDR notes, meeting link.

Status progression (demo simulation)

Status advances based on wall-clock time since the request was created — no background workers needed:

Elapsed Phase Confirmed
0–5s processing 0
5–15s matching 0
15s in_progress 1
+30s each in_progress +1
All seats filled completed N

So a 3-seat request is fully completed ~75 seconds after book_appointments. Tunable in src/tools/status.ts.


Setup

Requirements: Node.js 18+ (this repo was developed against Node 24).

npm install
npm run build

The compiled server entrypoint is dist/index.js. Data persists to data/store.json (auto-created).

Connect to Claude Desktop

Edit your Claude Desktop config file:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "leadclaw": {
      "command": "node",
      "args": ["C:\\absolute\\path\\to\\leadclaw-mcp\\dist\\index.js"]
    }
  }
}

Replace the path with the absolute path to your built dist/index.js. On macOS/Linux use forward slashes.

Restart Claude Desktop. The three LeadClaw tools should appear in the tool list.

Quick smoke test (without Claude Desktop)

node dist/index.js

Then paste the following on stdin (one JSON object per line):

{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"smoke","version":"1.0"}}}
{"jsonrpc":"2.0","method":"notifications/initialized"}
{"jsonrpc":"2.0","id":2,"method":"tools/list","params":{}}

You should see three tools advertised.


Project layout

leadclaw-mcp/
├── src/
│   ├── index.ts          MCP server entrypoint (stdio transport)
│   ├── tools/
│   │   ├── book.ts       book_appointments
│   │   ├── status.ts     check_status (time-based phase simulation)
│   │   └── details.ts    get_appointment_details
│   ├── mock/
│   │   ├── companies.ts  fake AI/SaaS company pool
│   │   ├── people.ts     fake names + default role pool
│   │   └── generator.ts  appointment + BANT + meeting-time synthesis
│   └── db/
│       └── store.ts      JSON file persistence
├── data/
│   └── store.json        auto-generated on first request (gitignored)
└── dist/                 build output

Demo recording flow

  1. Restart Claude Desktop with the MCP config in place.
  2. In a fresh chat, paste:

    Book 3 sales appointments next Tuesday in Dallas with AI startup CEOs. Budget $450.

  3. Claude calls book_appointments → you see the tool call card.
  4. Ask: "check status" → Claude calls check_status. First call (within ~15s) shows processing/matching; wait ~75s and call again to see all 3 confirmed.
  5. Optional: "give me the full briefing on the first one" → Claude calls get_appointment_details.

For the 30-second video cut, edit out the wait between calls.


What's intentionally not in this build

  • Real SDR matching / phone integration
  • Stripe / billing
  • OAuth (stdio is local-only)
  • Web dashboard
  • Anthropic Connectors Directory submission

These come after the concept lands.

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选