🌍 country-analytics-mcp

Ask an AI plain-language questions about world economics — answered live from three public APIs that disagree with each other.

"What was France's GDP in 2022 in US dollars?" · "Convert that to euros at the Dec 30 2022 rate." · "What's Europe's total GDP — and who did you exclude, and why?"

GDP, population, GDP per capita, regional totals in your currency, rankings — reconciled live (nothing hard-coded, no keys, no accounts) and fully traceable: every answer names its countries, years, FX rate, and exclusions.

⚡ Step 1 — Connect it to your AI (60 seconds, nothing to install)

The server is already running on Azure. Paste this URL into your AI client:

https://country-analytics-mcp.grayriver-35df865c.eastus.azurecontainerapps.io/mcp

🟠 Claude (claude.ai or Claude Desktop)

1. Settings → Connectors → Add custom connector

2. Name: country-analytics · URL: the one above · no authentication

3. Add, then start a chat and ask away — Claude picks the tools itself.

🟢 ChatGPT

1. Settings → Apps & Connectors → Advanced settings → enable Developer mode

2. Back in Connectors → Create: name country-analytics, MCP server URL as above, Authentication: None → save

3. In a new chat, enable the connector (via the composer's tools / Developer mode menu) and ask your question.

(Custom connectors require a paid plan on both platforms; exact menu names shift occasionally.)

🔍 No paid plan? MCP Inspector works too

npx @modelcontextprotocol/inspector

Transport Streamable HTTP → paste the URL → Connect → call the 8 tools from a form UI.

💤 The app scales to zero when idle — the first request after a quiet spell takes ~10 extra seconds to cold-start. Health check: /health

Related MCP server: World Bank MCP Server

💬 Step 2 — Things worth asking

💻 Run it locally (clean checkout → connected client)

No hosted dependency required — the same server runs on your machine with Python 3.10+:

git clone https://github.com/Satyampatel779/country-analytics-mcp.git
cd country-analytics-mcp
python -m venv .venv
.venv\Scripts\pip install .            # macOS/Linux: .venv/bin/pip install .

Claude Desktop (stdio): merge client-configs/claude_desktop_config.json into your claude_desktop_config.json (Settings → Developer → Edit Config), fix the absolute path, restart:

{
  "mcpServers": {
    "country-analytics": {
      "command": "C:\\path\\to\\country-analytics-mcp\\.venv\\Scripts\\python.exe",
      "args": ["-m", "country_analytics.server"]
    }
  }
}

MCP Inspector (stdio or HTTP): step-by-step in client-configs/mcp_inspector.md.

✅ Prove it works

.venv\Scripts\pip install ".[dev]"
.venv\Scripts\python -m pytest -q       # 11 live tests, both MCP transports

Expected values are cross-checked against independent direct fetches of the same live APIs — not constants.

🔧 What's under the hood

Three keyless public services, each speaking a different protocol, joined on ISO2 codes (never names — the sources spell them differently):

The 8 tools

Judgment calls are parameters, not hard-coded

• scope — geographic continent (default) vs. the World Bank region: for Europe the WB set adds Central Asia and drops Malta (the World Bank files it under MENA — flagged, of course).

• year — pinned exactly, or latest walks back past null years per country and reports the year used.

• missing_policy — exclude-and-list (default) or walk back to the most recent earlier value.

• fx_date — defaults to the pinned year's last trading day (2022 → 2022-12-30); pin any date you like.

• Refuses to guess — currencies outside Frankfurter's 30 (Serbian dinar, hryvnia…) return the supported list instead of an invented rate.

📄 More

• WRITEUP.md — sources & join choices, every disagreement found, every assumption, production notes (≤1 page, written for a non-technical stakeholder).

• samples/sample_run_output.md — a real timestamped run: the reconciled Europe table, computed metrics, and the flagged-issues list, kept as a reference in case a source is down during evaluation.