A workflow that it triggered automatically once per week (or even day) should be implemented to run the differential tests from protocol-envio/test and merkle-envio/test.
Some caveats to consider:
If the indexer reaches a block faster than the subgraph, the test will fail - this is somewhat of a false-positive as indexing speed is a good thing, so tests should be adapted for this caveat
The tests could be run against two separate subgraphs (different chain) for even more reassurance
Separately, availability tests could be considered for all subgraphs and indexers
A workflow that it triggered automatically once per week (or even day) should be implemented to run the differential tests from
protocol-envio/testandmerkle-envio/test.Some caveats to consider: