CKAN is currently the de-facto standard for these kind of portals, so being able to query the CKAN Action API enables to gather metadata from portals all over the world.
Some portals even have an API directly to their primary data (e.g. exposing a CSV as a querable JSON REST API), in CKAN this is called "DataStore". And this kind of data might be something you want to integrate. I can think of a number of datasets that could be interesting, e.g. schedule of garbage collection, public transport timetable or measurements from public weather stations.
Different portals have different data available. I'm from Switzerland and I know these datasets are available here. Is there a geograhical region you want to cover/not cover or it doesn't matter?