When the remote Dagger server is unreachable, setup_dagger_remote.sh
previously exited 0 with a misleading "CI will use the local Dagger
engine" message even when Docker was not running on the host.
This caused a confusing failure in the next step with a docker.sock
"no such file or directory" error rather than a clear early failure.
Add a `docker info` check after all remote probe attempts fail:
if local Docker is also unavailable, exit 1 with an actionable
error message so the root cause is immediately obvious.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
