We use GATK's UnifiedGenotyper in our pipeline:
http://www.broadinstitute.org/gsa/wiki/index.php/Unified_genotyper
It can take a dbSNP VCF file, like the dbSNP 132 release:
ftp://ftp.ncbi.nlm.nih.gov/snp/organisms/human_9606/VCF/v4.0/00-All.vcf.gz
and assigns rs associations as the 3rd ID column of the VCF file.
Exactly the sort of info I was looking for. I'd not known of Unified genotyper but am pleased to see open code and a wiki manual. Happily Promethease already knows how to read VCF
Hat tip to Daniel MacArthur for drawing my attention to the format early on.
The Galaxy folks likely have some good ideas about integration,
but one approach could be to have a tool or display app that takes
a VCF files with human genome builds and produces HTML with links
to resources such as SNPedia. This would just need linkouts like
http://snpedia.com/index.php/rs1234.
Once you have an rs# assigned, guessing the SNPedia page name isn't hard enough to justify any code. There would be value if it can also resolve the strand.
I'm thinking about a SNPedia hosted service for running Promethease directly from galaxy, and returning a suitably structured response for further processing. But it's very early in my thinking.