MCP Guide

Talk to the data. Your AI does the rest.

Connect Claude, Cursor, Gemini, or ChatGPT to the ElkassabgiData libraries once, then just ask in plain English. Your assistant finds the exact series, downloads it, and hands it back with the citation and license attached — across 1-minute US equities and billions of global economic series.

New to MCP? Read this first.

MCP (Model Context Protocol) is a small, open standard that lets an AI assistant plug into an outside service — the same way a USB-C port lets any device plug into any charger. You connect the ElkassabgiData MCP server to your AI once, and from then on the AI can search our catalog, pull data, and read the licensing rules on its own, right inside your conversation.

You don't write code to set it up. You paste one URL (or a short config) into your AI tool. After that you type questions like a normal chat — "how many people live below the poverty line in Poland?" — and the assistant does the librarian work: it figures out which dataset answers that, downloads it, and cites it.

Plain English in

No series codes to memorize. Ask the way you'd ask a research assistant. The AI translates your question into the right dataset.

Citations out

Every series comes back with its source, license, and attribution line already attached — ready to drop into a paper.

One key, every library

Your free ElkassabgiData key unlocks both the HF equity data and the economic data through the same connection.


What it actually looks like

These are real series the connected assistant finds live — not mock-ups. The AI runs a search, reads the titles, and picks the one that fits (asking you to disambiguate when a question has more than one correct answer). Then it downloads and cites it.

Poverty in Poland Econ Data Library
How many people live below the poverty line in Poland?
→ the AI calls search_econ_series("poverty Poland")
I found several official measures — "below the poverty line" can mean different thresholds, so here are the main ones:
worldbank_wdi:SI.POV.NAHC Poverty headcount ratio at national poverty lines (% of population)
pip:PIP:headcount:POL:215 Poverty headcount ratio at $2.15/day (2017 PPP) — Poland
pip:PIP:headcount:POL:1000 Poverty headcount ratio at $10.00/day (2017 PPP) — Poland
Which line do you want — Poland's national definition, or an international benchmark?
The national one.
get_econ_series("worldbank_wdi:SI.POV.NAHC"), filtered to Poland
Here's the series as a table, plus the World Bank citation and license to include if you publish. Want it charted?
Trade into Australia Econ Data Library
How much does Australia import each year?
search_econ_series("Australia imports total trade")
The headline annual import value:
unctad_gasbtoia:…imports…total-trade-in-goods-and-services.australia Imports — Total trade in goods and services — Australia (US$ at current prices, millions)
Want goods-only instead of goods + services, or the trade balance alongside it?
Goods and services is fine. Download it.
get_econ_series(…) — returns rows + UNCTAD citation
Done — annual series delivered with its UNCTAD attribution line attached.
Intraday event study HF Data Library
Pull AAPL 1-minute bars and show me what happened intraday around 2024-08-05.
get_hf_download_link("AAPL"), then the AI runs the analysis in code
Downloaded the clean 1-minute file, sliced to that week, and computed minute-by-minute returns and realized volatility around the drop. Here's the chart and the numbers — with the dataset DOI to cite.

The intelligence is your AI plus honest metadata working together: it searches, reads real titles and units, and never invents a number — a wrong guess returns an honest "not found," never fake data. See the full worked prompts on the AI Tools page.


Set it up — pick your AI tool

You need a free ElkassabgiData API key (looks like hfd_…). Browsing and searching work without one; downloading data needs it. Sign in and every command below fills your real key in automatically.

Not signed in. The snippets below show YOUR_KEY as a placeholder. Sign in or get a free key and this page will drop your real key into every command.

Claude — web, desktop & mobile

Works on Free, Pro, Max, Team & Enterprise. (Free plans allow one custom connector.) No install — it's all in Settings.

  1. Open Settings → Connectors (in the Claude app or on claude.ai).
  2. Click Add custom connector.
  3. Name it ElkassabgiData and paste this URL (your key is embedded, so there's nothing else to configure):
    https://elkassabgidata-mcp.elkassabgi.workers.dev/mcp?api_key=YOUR_KEY
  4. Click Add, then start a chat and ask a data question. Claude will show the ElkassabgiData tools it's using.

Claude connects to the server from Anthropic's cloud (not your laptop), which is why a hosted URL like ours works with no local setup.

Claude Code (terminal)

One command in your shell. Free to install: npm i -g @anthropic-ai/claude-code.

  1. Add the server over HTTP transport:
    claude mcp add --transport http elkassabgidata \
      "https://elkassabgidata-mcp.elkassabgi.workers.dev/mcp?api_key=YOUR_KEY"
  2. Prefer a header to keep the key out of the URL? Use this form instead:
    claude mcp add --transport http elkassabgidata \
      https://elkassabgidata-mcp.elkassabgi.workers.dev/mcp \
      --header "Authorization: Bearer YOUR_KEY"
  3. Run claude and type /mcp to confirm elkassabgidata is connected, then ask away.

Cursor

Add a global server in ~/.cursor/mcp.json (or use Settings → Tools & MCP → New MCP Server).

  1. Open (or create) ~/.cursor/mcp.json and add:
    {
      "mcpServers": {
        "elkassabgidata": {
          "url": "https://elkassabgidata-mcp.elkassabgi.workers.dev/mcp",
          "headers": { "Authorization": "Bearer YOUR_KEY" }
        }
      }
    }
  2. Save, then open Settings → Tools & MCP and confirm elkassabgidata shows a green dot.
  3. In chat (Agent mode), ask a data question — Cursor will call the tools.

Gemini CLI

Configure in ~/.gemini/settings.json. Note Gemini uses httpUrl (not url) for remote servers.

  1. Add the server to ~/.gemini/settings.json:
    {
      "mcpServers": {
        "elkassabgidata": {
          "httpUrl": "https://elkassabgidata-mcp.elkassabgi.workers.dev/mcp",
          "headers": { "Authorization": "Bearer YOUR_KEY" }
        }
      }
    }
  2. Start gemini and run /mcp to verify the server and its tools loaded.
  3. Ask your question in the CLI.

ChatGPT

Custom MCP connectors live in Developer Mode (web). Available on Plus, Pro, Business, Enterprise & Edu; ChatGPT supports remote HTTPS servers like ours.

  1. Go to Settings → Connectors → Advanced and turn on Developer Mode.
  2. Back in Connectors, add a new remote MCP server / app with this URL:
    https://elkassabgidata-mcp.elkassabgi.workers.dev/mcp?api_key=YOUR_KEY
  3. Save and refresh the connector so it pulls in the tools. In a chat, enable it from the Developer Mode / tools menu and ask your question.

A quick correction on a common mix-up: ChatGPT is OpenAI's product (not built on Claude), but it speaks the same open MCP standard, so this one server works for Claude, ChatGPT, Gemini, and Cursor alike.

Your connect URL and configs contain your real key — treat them like a password. Don't paste them into shared or public chats or commit them to a public repo. If a key leaks, regenerate it from your account page; the old one revokes instantly.

The tools your AI gets

Once connected, the assistant can call any of these. Search and metadata are open; only actual data downloads use your key.

ToolWhat it doesKey?
search_econ_seriesSearch billions of economic series across 300+ sources
get_econ_seriesFetch a series as rows, with citation + license inlinerequired
get_econ_series_metadataFull metadata: license, attribution, coverage
list_econ_sourcesEvery source with its license
get_data_freshnessLive per-source update board (honest staleness flags)
get_hf_download_linkAuthenticated links for 1-min bars / 25 academic variablesfor use
get_hf_variables_dictionaryThe 25 variables' exact definitions
get_family_statusLive stats + data currency for both libraries
get_auth_statusWhether a key is configured on this connection (masked)

The server also publishes MCP resources — a data-honesty charter (survivorship, the 2022 HF source break, per-source licensing, freshness semantics) plus the variables dictionary — and guided prompts (analyze_econ_series, compare_countries, hf_event_study) that bake the honesty rules into each workflow.

Fair use & licenses. MCP traffic uses the same free account, the same 100 downloads/minute limit, and the same per-series licenses as the website. Each response carries its own license — please honor them (especially non-commercial sources) and keep the attribution lines in published work. Your key is used server-side only; it is never echoed into the AI conversation.

Troubleshooting

The AI says it can't reach the server.
Confirm the URL is exactly https://elkassabgidata-mcp.elkassabgi.workers.dev/mcp. In Claude Code and Gemini CLI, run /mcp to see the connection status.
Search works, but downloads fail with an auth error.
Your key is missing or wrong. Sign in above so the snippets fill your real key, or copy it from your account page. Search and metadata are keyless by design; only get_econ_series and HF downloads need the key.
It found the wrong series.
Add detail to your question (units, country, time period, "national vs international poverty line"). The assistant re-searches with better terms — that's the normal loop, not a failure.
Which key do I use — HF or Econ?
There's only one. Your HF Data Library key is your ElkassabgiData key; it works across every library through this one server.
Can I use it with no key at all?
Yes, partially. Searching the catalog, reading metadata, listing sources, and checking freshness all work keyless. Only pulling the actual data requires the free key.

Prefer copy-paste prompts over connecting a server? The AI Tools page has ready-made prompts for downloading and analyzing the data. Questions or a setup that won't cooperate? Email me.