ADOxx uses a relational database, which can be accessed via ADOxx.
Hence, models that either have been modeled manually using the graphical modeling toolkit, or that have been generated using software by invoking appropriate interfaces are stored in a repository.

Those models can be accessed either by a special query language, or can be accessed by an internal script language called “ADOScript”.
So there is a query engine that enable the query the repository either by:
(a) a simple user interface in the modeling toolkit,
(b) a pre-defined query that is constructed as part of the meta model or
(c) an implementation in the script language.

Query samples are introduced either in the tutorial slides set: http://www.adoxx.org/live/documents/10157/d228d6f1-309d-45a4-a9b8-aa914a0634af
 /slides/ file “Mechanisms and Algorithms” slide 13pp for the simple user interface (a), slide 23pp for the pre-defined queries (b) and 98 pp for implemented queries (c).

There is also an interesting sample on queries in the tutorial set:
http://www.adoxx.org/live/documents/10157/28257/2.+Implementing+an+Algorithm.zip/5c816eed-a844-4d6c-90d6-6ea8b45b7155

So the answer is: ADOxx has a repository storing models and provides a query engine enabling to query those data. Those queries can be manually triggered, or implemented in a script language.