DEV Community

firstdata
firstdata

Posted on

FirstData: Open Knowledge Base of 160+ Authoritative Global Data Sources with MCP Integration

#opensource #ai #mcp #datasets

The Problem

AI hallucination is one of the biggest challenges with LLMs today. When your AI agent confidently cites statistics, regulations, or market data — where does that data actually come from?

The solution isn't just bigger models. It's giving AI agents access to verified, authoritative primary sources.

What is FirstData?

FirstData is an open-source knowledge base of 160+ curated authoritative data sources from around the world, with a built-in MCP (Model Context Protocol) server for AI-powered discovery.

Think of it as a "trusted data directory" for AI agents.

What's Inside

  • 🏛️ 59 Government sources — US Census, Federal Reserve, China NBS, Eurostat
  • 🌐 44 International organizations — World Bank, WHO, IMF, WTO, FAO
  • 🔬 28 Research institutions — NBER, CEPR, major universities
  • 📈 14 Market data providers — Bloomberg, LSEG
  • 🗺️ 50+ domains across economics, health, environment, education
  • 🌍 Global coverage — 67 global, 69 national, 20 regional sources

MCP Integration

FirstData provides a standard MCP server that any MCP-compatible AI agent can connect to:

{
  "mcpServers": {
    "firstdata": {
      "url": "https://firstdata.deepminer.com.cn/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_TOKEN"
      }
    }
  }
}
Enter fullscreen mode Exit fullscreen mode

6 tools available: search sources, get details, browse by domain/country, get statistics, and submit feedback.

Why It Matters

  1. Combat AI Hallucination — Point your agents to real, verified data sources
  2. Save Research Time — Stop googling for "where to find GDP data" — just ask your AI agent
  3. Structured Metadata — Every source has standardized JSON with descriptions, API URLs, update frequency, and coverage
  4. Bilingual — Full English and Chinese descriptions for every data source
  5. Open Source — MIT licensed, community-driven, continuously updated

Use Cases

  • Market Research — Find authoritative industry data sources instantly
  • AI Agent Development — Give your agents trusted data discovery capabilities
  • Academic Research — Discover core datasets in specialized fields
  • Fact-Checking — Trace information back to its authoritative origin

Get Started

  1. ⭐ Star the repo: github.com/MLT-OSS/FirstData
  2. Apply for API access: firstdata.deepminer.com.cn
  3. Connect via MCP: Add to Claude Desktop, Cursor, or any MCP client
  4. Contribute: Add new data sources via PR — we welcome contributions!

If you find this useful, please ⭐ Star the repo! Contributions welcome.

Open-sourced by Mininglamp Technology (2718.HK)

Top comments (0)