Automat ingests many sources. For each source it stands up an API. e.g. there is automat-ctd, automat-hmdb, etc.
Then, a subset of these are merged into robokop, which is also placed into automat. Strider finds both ctd and robokop in smartapi, and queries both, leading to dupe edges and inefficiency.
Handling https://github.com/ranking-agent/strider/issues/394 will solve the problem of duplicate edges, but do we want to have some special handling to not call robokop or other aggregate graphs in automat?
Automat ingests many sources. For each source it stands up an API. e.g. there is automat-ctd, automat-hmdb, etc.
Then, a subset of these are merged into robokop, which is also placed into automat. Strider finds both ctd and robokop in smartapi, and queries both, leading to dupe edges and inefficiency.
Handling https://github.com/ranking-agent/strider/issues/394 will solve the problem of duplicate edges, but do we want to have some special handling to not call robokop or other aggregate graphs in automat?