Are there any guidelines on data packages in bioconductor. I am interested in contributing a curated data package from specific TCGA dataset. I would like to know how the data should be formatted and how big the data can be in the package. 

Generally we would encourage making a data package that uses ExperimentHub -

Before you proceed, please see the curatedTCGAData package that already exists in Bioconductor and includes/provides data from TCGA.