Consider caching org and project IDs

CBShell stores as little state as possible, since things get sticky once we start doing that. However commands against Capella clusters are running a bit slow because we are finding the org and project ID every time. We should consider setting them as part of the state when we fetch them for the first time, and read from here on future commands.

Start by tracing how CBShell's Capella cluster commands find the organization and project IDs, then inspect where command state is stored. The work is done when IDs fetched for the first time can be reused by later commands without changing command behavior; verify that repeated commands no longer perform the same lookups.

Written by the indexing model from the issue text.