/links is POST only, and /items/{id}/links is per item, so building a dependency graph across a workspace costs one API call per item.
Something like GET /links?workspace_id=X would make it a single call.
For context, we compute "what should I work on next" from the blocking graph, which means reading every link between unfinished items on each run.
CQL cannot filter on links, and /items/changes does not report link mutations, so there is no cheaper path and no safe way to cache it.
/linksis POST only, and/items/{id}/linksis per item, so building a dependency graph across a workspace costs one API call per item.Something like
GET /links?workspace_id=Xwould make it a single call.For context, we compute "what should I work on next" from the blocking graph, which means reading every link between unfinished items on each run.
CQL cannot filter on links, and
/items/changesdoes not report link mutations, so there is no cheaper path and no safe way to cache it.