Use case rules. Imagine a rule that triggers on any insert: * of a language literal in German and uses an LLM to translate it automatically into French and Italian and store these in the store. * of a literal with a `DNA` datatype immediately insert the reverse complement as well as biological use case. * of a literal with a `value in grams` datatype automatically insert an equivalent value in troy ounces, etc. A longer term goal is for "virtual triples" or index maintenance. Mostly around use cases raised in comments to this [SPARQL-DEV](https://github.com/w3c/sparql-dev/issues/129).