currently we specify images one by one in he front matter... this should be automated
currently we specify images one by one in he front matter... this should be automated