Impact: Teams will see deployment failures or, worse, see successful pipeline runs that do not act on all applicable targets. This is worse because they will see success but it did not actually succeed completely.

Workaround: Similar to above, we can call into Harness pipelines from code we maintain outside of Harness, passing in the appropriate environments.

Request: We know Harness supports a static inventory of environment targets. We need to query an API on each pipeline run to get the current status of target clusters and allow teams to define or skip which of these targets should be used via tags or labels. The inventory can change with every pipeline run.