Full SpamAssassin breakdown of NON-spam messages?

[Jeremy Morton]

OK, so SA gives you a breakdown in the msg body of why it thought 
messages tagged as spam were spam.  But what about false negatives? 
It'd be really useful if SA could give me a full breakdown of its 
analysis in the e-mail's header, some kind of 'debug mode', so I could 
tell why it thought a message was not spam.  For example, a recent spam 
message got through and this was in the header:

X-Spam-Status: No, score=-2.3

I want way more detail than that.  Is there a way to get SA to put a 
full analysis in the header, including rules matched and their scores?

Best regards,
Jeremy Morton (Jez)