Golddigger MCP Server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Golddigger MCP Serversearch for AI startups in Berlin and enrich the top result"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Golddigger MCP Server
B2B Lead Intelligence API — search, enrich & score 10M+ global companies from any MCP-compatible client.
What is this?
A Model Context Protocol (MCP) server that wraps the Golddigger API — giving AI assistants (Claude, Cline, Cursor, etc.) the ability to search, enrich, and score B2B leads in real time.
4 tools exposed:
Tool | Description | Data Flow |
| Service health + real account quota ( | External API call |
| 挖客前预检:服务端剩余 + 本地每日预算账本 | External API call |
| Search by buyer-intent keyword + country ( | External API call, 1 quota each |
| Six-dimension quality scoring | Local only |
| Full radar chart + Word report | Local only |
| Role-aware outreach draft + local queue CSV | Local only, 0 quota |
Related MCP server: Prospeo MCP Server
Data Sovereignty
✅ Report generation: 100% local (matplotlib + python-docx). No data leaves the user's machine.
✅ Scoring: Pure local computation — six-dimension model runs on-device.
⚠️ Search: Calls Golddigger API. Customer provides their own API key — costs borne by customer.
Installation
git clone https://github.com/SassumiX/golddigger-mcp.git
cd golddigger-mcp
pip install -r requirements.txtConfiguration
# Set your Golddigger API key (customer provides their own)
export GOLDDIGGER_API_KEY="gd_your_key_here"
export GOLDDIGGER_HOST="https://golddigger.gold" # 可选
export GOLDDIGGER_DAILY_BUDGET=20 # 可选:本地软预算,默认 20
export GOLDDIGGER_QUOTA_RESERVE=2 # 可选:剩余低于此值不再自动搜Token billing: The
search_leadstool uses the customer's API key. Report generation is free (no external API calls).
Usage
Claude Desktop (macOS)
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"golddigger": {
"command": "python3",
"args": ["/FULL/PATH/TO/golddigger-mcp/server.py", "--transport", "stdio"]
}
}
}Restart Claude Desktop.
Cline / Other MCP Clients
{
"mcpServers": {
"golddigger": {
"command": "python3",
"args": ["/FULL/PATH/TO/golddigger-mcp/server.py", "--transport", "stdio"],
"env": {
"GOLDDIGGER_API_KEY": "gd_your_key_here"
}
}
}
}HTTP Endpoint
python3 server.py --transport streamable-http --port 8080MCP Tools
generate_report(商业核心工具)
用户说:"出客户画像报告"
Agent 自动调用 → 生成本地 .docx 文件# Agent 调用示例
result = golddigger.generate_report(
company_name="cnpulk.com",
product_desc="DTH drill bits, rock drilling tools",
buyers=[
{
"name": "Afrimat Limited",
"country": "南非",
"industry": "Mining",
"has_contact": True,
"scale": "上市矿业集团",
"buyer_role": "Contract Mining总监",
"purchase_needs": "DTH钻头,持续消耗",
"contacts": ["pierre@afrimat.co.za · +27 82 411 7475"],
"outreach": "We offer 20-30% cost savings on DTH bits."
},
# ... 2 more buyers
],
api_key="gd_customer_own_key" # 客户自己的Key,费用自担
)
# result["output_path"] → 本地 .docx 文件路径score_lead
result = golddigger.score_lead(
company="Afrimat Limited",
country="南非",
industry="Mining",
has_contact=True,
scale="上市矿业集团",
note="并购整合期"
)
# result["total"] = 91, result["tier"] = "HOT"Pricing
Plan | 每日配额 |
Free | 20 次/天(2026-10-06 由 |
Paid |
⚠️ 别信任何写死的配额数字,以
check_quota的实时返回为准。 Get your key at https://golddigger.gold
outreach 是主线(不是 search)
search 每天只有 20 次,draft_outreach 不限次数、不耗配额 —— 因为找人的成本高,
开口的成本为零。所以工作流必须是:
check_quota → search_leads(dry_run) → search_leads → 对每条 lead 跑 draft_outreach → 进 outreach_queue.csv → 按 D0/D3/D7/D14 跟进draft_outreach 会按公司描述自动识别角色(进口商 / 地产开发商 / 酒店 / 民宿)并切换话术,
支持 email / linkedin / whatsapp / wechat 与中英双语,队列按 company+channel 去重。
r = golddigger.draft_outreach(
lead={"company": "Bin Harkil Group", "contact": "Procurement",
"country": "SA", "industry": "real estate developer"},
my_profile="smart lock factory, OEM/ODM, CE, MOQ 200, 7-day samples",
channel="email", language="en",
)
# r["subject"], r["body"], r["followup_plan"], r["quota_cost"] == 0已知问题 / 端点实测(2026-10-06)
端点 | 状态 |
| ✅ |
| ✅ 真实配额端点(旧代码用的 |
| ✅ |
| ⚠️ 当前 500,站点侧故障,重试无用(5xx 不扣配额) |
| ❌ 从未部署(旧 |
配额纪律(Agent 必读)
批量挖客前先调
check_quota,别让 Agent 自由发挥地循环搜索。search_leads(dry_run=True)可以预演请求,不烧配额 —— 先批量预演,再挑着真跑。本地预算账本写在
~/.golddigger/search_budget.json,默认上限 20/天,按天重置。剩余 ≤ 2 次(
GOLDDIGGER_QUOTA_RESERVE)时search_leads直接拒绝发起,留额度给人工决策。优先级:采购角色词 > 产品细分词 > 泛需求词。泛需求词最费额度、命中率最低。
License
MIT — SassumiX 2026
This server cannot be deployed
Maintenance
Related MCP Connectors
Lead enrichment for AI agents: email finder, company enrichment, people data, firmographics.
Search B2B people, companies and local businesses and build lead lists from your AI assistant.
AI-native B2B sales research, ranking, and CRM enrichment.
Real-time B2B data for agents: search and enrich 1B+ people and 200M+ company profiles.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables AI assistants to automate sales prospecting by finding contacts by role and industry, enriching data with emails and tech stacks, scoring against ideal customer profiles, and generating personalized outreach sequences. Streamlines lead generation and sales engagement workflows through integrated research and sequence generation tools.MIT

Prospeo MCP Serverofficial
AlicenseAqualityDmaintenanceEnables AI tools to search and enrich B2B leads, including finding professional emails, company profiles, and filtering people and companies by various criteria.5283 npmMIT- AlicenseAqualityDmaintenanceEnables AI agents to research companies and find contacts with structured data from multiple free sources, including company info, tech stack, and email addresses.3MIT
- AlicenseAqualityDmaintenanceEnables AI assistants to search and retrieve B2B leads and company data from the LeadBrew database, including contact details and company information.51MIT