Bayesian clustering with variable selection is a fully automatic clustering method, designed for low-sample-size-high-dimensional situations. This might be regarded as a high-dimensional variant of the mclust package.
The project summary page you can find here.
The project description and documentation page you can find here.