RocketReach MCP Worker
MCP server for RocketReach, enabling people and company lookup via search, full profile, and company details through Cloudflare Workers.
README
RocketReach MCP — Cloudflare Worker
A remote Model Context Protocol server for the RocketReach API, deployed as a Cloudflare Worker.
Exposes three tools that any MCP-compatible client (Claude Desktop, Cursor, Windsurf, etc.) can call:
| Tool | Description |
|---|---|
search_people |
Search people by name, employer, title, or location |
lookup_person |
Full profile + verified contact info (emails, phones) |
lookup_company |
Company details: size, revenue, industry, location |
Why a Cloudflare Worker?
The server uses createMcpHandler from the Cloudflare Agents SDK — a stateless MCP handler that runs in a plain Worker (no Durable Objects needed). Each request is fully self-contained; the RocketReach API key is read directly from the incoming ROCKETREACH_API_KEY header, so multiple users can share the same deployment with different keys.
Quick deploy
# 1. Clone and install
git clone https://github.com/balazsnemeth/rocketreach-mcp-worker.git
cd rocketreach-mcp-worker
npm install
# 2. Log in to Cloudflare (first time only)
npx wrangler login
# 3. Deploy
npm run deploy
The worker is now live at https://rocketreach-mcp.<your-subdomain>.workers.dev/mcp.
Optional: store a default API key as a secret
If you want the worker to work without a per-request header (e.g. for personal use), store your key as a Worker secret:
npx wrangler secret put ROCKETREACH_API_KEY
Client configuration
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):
{
"mcpServers": {
"rocketreach": {
"command": "npx",
"args": [
"mcp-remote",
"https://rocketreach-mcp.<your-subdomain>.workers.dev/mcp",
"--header",
"ROCKETREACH_API_KEY:<your-rocketreach-api-key>"
]
}
}
}
Cursor / Windsurf
Add to your MCP settings file:
{
"rocketreach": {
"command": "npx",
"args": [
"mcp-remote",
"https://rocketreach-mcp.<your-subdomain>.workers.dev/mcp",
"--header",
"ROCKETREACH_API_KEY:<your-rocketreach-api-key>"
]
}
}
Note:
mcp-remoteproxies the local stdio transport (used by MCP clients) to the remote HTTP endpoint. Install it globally if needed:npm install -g mcp-remote.
Tools reference
search_people
Search for people in RocketReach. Returns basic profile information — no contact details are consumed from your plan's credit quota.
Parameters
| Parameter | Type | Description |
|---|---|---|
name |
string | string[] |
Full name or list of names — e.g. "Elon Musk" |
current_employer |
string | string[] |
Company name or list — e.g. "Tesla" |
title |
string | string[] |
Job title or list — e.g. "CEO" |
location |
string | string[] |
City / region or list — e.g. "San Francisco" |
radius_miles |
number |
Radius around location in miles (optional) |
start |
number |
1-based pagination offset (default: 1) |
page_size |
number |
Results per page, max 100 (default: 20) |
At least one of name, current_employer, title, or location is required.
Response shape
{
"total_results": 42,
"start": 1,
"next": 21,
"profiles": [
{
"id": 12345678,
"name": "Elon Musk",
"current_employer": "Tesla",
"current_title": "CEO",
"location": "Austin, TX",
"linkedin_url": "https://www.linkedin.com/in/elonmusk"
}
]
}
lookup_person
Get a full person profile including verified contact information. Uses one credit per unique lookup.
Parameters
| Parameter | Type | Description |
|---|---|---|
identifier |
string |
The identifier value (see lookup_type) |
lookup_type |
string |
How to identify the person (default: "id") |
lookup_type values
| Value | identifier format |
Notes |
|---|---|---|
id |
RocketReach numeric ID | Fastest, most reliable |
linkedin_url |
Full LinkedIn profile URL | |
email |
Email address | |
name_employer |
"Full Name|Company Name" |
Pipe-separated |
npi |
NPI number | Healthcare providers only |
Response includes id, name, current_employer, current_title, location, linkedin_url, status, contact_info (emails, phones, work email), job_history, education, skills, and social_profiles.
lookup_company
Get detailed company information.
Parameters
| Parameter | Type | Description |
|---|---|---|
identifier |
string |
The identifier value (see lookup_type) |
lookup_type |
string |
How to identify the company (default: "domain") |
lookup_type values
| Value | Example |
|---|---|
domain |
"rocketreach.co" (preferred) |
name |
"RocketReach" |
linkedin_url |
Full LinkedIn company URL |
ticker |
"AAPL" |
id |
RocketReach internal company ID |
Response includes id, name, domain, website, linkedin_url, description, industry, employees, employee_range, revenue, revenue_range, city, state, country, address, founded_year, company_type, ticker, and logo_url.
Local development
npm run dev
The worker is served at http://localhost:8787. Point mcp-remote at it for local testing:
{
"rocketreach": {
"command": "npx",
"args": [
"mcp-remote",
"http://localhost:8787/mcp",
"--header",
"ROCKETREACH_API_KEY:<your-key>"
]
}
}
API key priority
ROCKETREACH_API_KEYrequest header (sent bymcp-remote --header) — per-user, per-requestROCKETREACH_API_KEYWorker secret — shared fallback (set viawrangler secret put)
If neither is present the worker returns 401.
Tech stack
- Cloudflare Workers — edge runtime
- Cloudflare Agents SDK —
createMcpHandlerfor stateless MCP transport - @modelcontextprotocol/sdk — MCP protocol
- Zod — tool input validation
- TypeScript
License
MIT
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。