MCP

italia-mcp-servers: the right way to use this catalog of Italian MCP servers and get started for the first time

If you want to quickly find ready-made MCP servers for Italian public data, PA, or e-invoicing scenarios, start with this curated 30-project catalog for shortlisting; but note it is only an index—actual installation and usage require jumping to each individual server project to confirm its requirements.

Official source · bsab/italia-mcp-servers

Install

curl -s https://bsab.github.io/italia-mcp-servers/catalog.json | jq '.servers[] | .name'

This is not an MCP server you can install directly, but a curated catalog of Italian MCP servers: it lists 30 projects covering public data, legal-tech, e-invoicing, public administration, and related fields—16 implemented in Python, 13 in TypeScript, and 1 in JavaScript. Its core value is helping you quickly find ready-made MCP servers built for Italian use cases. The most important thing to know before using it: the catalog itself provides no services—you must follow each entry to the corresponding project and install it separately, and each server has its own license and its own runtime requirements.

Who it's for

  • Users who need to integrate Italian public data, public administration (PA), e-invoicing, or legal-tech capabilities and want to compare which ready-made MCP servers exist before committing.
  • Developers who want to consume this catalog programmatically in bulk—an official, curl-fetchable JSON catalog file is provided.
  • Who it's NOT for: anyone looking for a single, out-of-the-box MCP server—this repository is just an index and contains no runnable server itself. If you want capabilities outside Italian domains or from a specific server, you must confirm them on that server's official page; the catalog evidence does not detail each server's installation steps.
  • Prerequisites: the catalog is published both as a web page and as JSON. The catalog itself is MIT-licensed, but each listed server retains its own license—check each one individually before adopting it.

First try in 3 minutes

The most direct official path is to fetch the machine-readable JSON catalog with a single command and browse what servers are available.

Prerequisites: curl locally (the official example also uses jq for formatted output) and access to GitHub Pages.

Run the official command in your terminal:

curl -s https://bsab.github.io/italia-mcp-servers/catalog.json | jq '.servers[] | .name'

This is a purely read-only network request: it only downloads the public catalog data and prints each server's name, writing nothing locally and producing no remote side effects.

Success signal: the terminal prints the list of server names included in the catalog, one per line. If jq is not installed, drop the pipe and just run curl -s https://bsab.github.io/italia-mcp-servers/catalog.json to view the raw JSON.

Next step: after picking interesting servers from the output and the catalog web page, you must visit each project's own page to see its actual installation and configuration instructions—this catalog's evidence does not provide installation commands for individual servers, so don't assume they all work the same way. If you want to add your own server to the catalog, the official path is to create a kebab-case-named JSON file in the servers/ directory, fill it in according to schema/server.schema.json, then validate with python3 scripts/validate_servers.py and rebuild the README with python3 scripts/build_readme.py (dependencies can be installed with python3 -m pip install -r scripts/requirements.txt); CI runs the same checks on pull requests.

Use cases

  1. Survey and shortlisting: you need ready-made MCP servers for an agent project that requires Italian public data or PA capabilities. Pull catalog.json as described in the main path above and list all server names, then narrow down candidates using the categories in the catalog web page (public data, legal-tech, e-invoicing, public administration, etc.—6 categories in total), and finally visit each candidate project's official page to confirm its runtime requirements and license.
  2. Automated integration: you want your toolchain to sync this catalog periodically. Use the same curl command to pull catalog.json into your script and parse the servers array in the JSON programmatically (the official example iterates .servers[] with jq). What can be confirmed at this step: the data source is the officially published JSON file; as for the exact schema field names of each field, defer to the actual contents of catalog.json—this evidence does not list them one by one.

Boundaries of use

  • Read/write nature of this repository: for users it is a purely read-only catalog (web page + JSON); only the contributor workflow involves write operations (submitting JSON files and running build scripts, with side effects limited to your own repository copy and pull requests, validated by CI).
  • Credentials and local data: fetching the catalog requires no credentials or keys; each listed server's requirements around credentials and local files depend on that server itself—this catalog's evidence does not cover them.
  • Clear limitations: the catalog contains 30 entries in total, focused mainly on Italy-related domains; licenses apply per entry.
  • Not yet confirmed: whether each server provides a readOnlyHint, their individual runtime requirements, client compatibility, and privacy boundaries are all outside the scope of this catalog's evidence. Before adopting any server you must verify these on its official page—the fact that a server is listed in the catalog must not be treated as an endorsement of its security or quality.