datagov-mcp

MCP.Pizza Chef: aviveldan

Thirteen tools sit on top of the Israeli government's open data portal: ten for finding datasets, reading what is in them and pulling out rows, plus three that draw an interactive summary table, a chart or a map. No key or account is needed, because the portal is public. Two things to expect. Dataset names and column headings are largely in Hebrew, and charts are built from the first hundred rows only, so a picture of a large dataset is a sample rather than the full story.

Data
Web/Research

Use This MCP server To

Search Israeli government data for a topic I care about See what columns and value ranges a dataset holds Chart which cities have the most housing lottery entrants Plot public transport stops on a map Pull the actual rows out of a public dataset Compare two datasets before deciding which to use

README

DataGov Israel MCP Server

An MCP server for exploring Israeli government open data (data.gov.il) — with built-in interactive visualizations powered by MCP Apps.

Search thousands of public datasets, profile their structure, generate charts, and plot geographic data on maps — all from your AI assistant.

Tests Python 3.10+ License: MIT


What Can You Do With This?

🏠 Explore the Housing Market

Profile public housing datasets to understand unit sizes, locations, and availability:

Dataset Profile — Public Housing

See which cities have the most demand in government housing lotteries:

Housing Lottery Subscribers by City

Understand the distribution of apartment sizes across the country:

Distribution of Public Housing Unit Sizes

Track housing unit availability over time:

Housing Units Available per Lottery

🗺️ Map Public Infrastructure

Visualize education institutions across Israel:

Education Institutions Map

Plot public transport stations:

Public Transport Stations


datagov-mcp FAQ

Do I need a key or an account?
No. The Israeli open data portal is free and public, so there is nothing to sign up for and nothing to pay.
Are the charts drawn from all the data?
No, and this matters. A chart uses the first hundred rows and a map the first five hundred, with no totalling up beforehand, so a picture of a large dataset shows a slice rather than the whole thing.
Which apps does it work in?
It installs into Claude Desktop with one command. The charts, tables and maps only appear in apps that support interactive panels; elsewhere you still get the underlying numbers.
Can I use this to explore the housing market?
Yes. Search for the housing lottery or public housing datasets, ask what fields they contain, then ask for a chart of entrants by city or a spread of apartment sizes.
Is the data in English?
Mostly not. Dataset titles and column names are largely Hebrew, so you will often be reading Hebrew field names even when the conversation is in English.
Do the map examples in the readme all work?
Not all. The schools dataset it suggests stores its coordinates in a different system and nothing converts them, so those points land in the wrong place. The transport stations dataset has plain latitude and longitude and maps correctly.
How hard is the setup?
Technical. You copy the code down, create a Python environment, install it and then run one command to register it with Claude Desktop.
Does anything leave my computer?
Yes. Every request goes to the government portal, and maps additionally load their mapping software and map tiles from public third-party sites.