User agents are how various pieces of Internet software like browsers identify themselves to a website. It's just some text that might, in my case, tell a website I'm running IE6 SP1 on Win2000. Or it might be Google telling me she's back for another dance. Or it might be some nefarious website stripper that's there to bring my server to its knees.
The user agent field in your logs is from the value of the HTTP_USER_AGENT field found in your visitors HTTP request header. This value is assigned by the software your visitor runs. This value is very easily and frequently spoofed. Even mildly sophisticated visitors can set the user agent field to anything they want.
Most agents identify themselves clearly and honestly. But, there are a lot of bad agents that mis-identify themselves for different purposes. For example, software robots that crawl the web looking for email addresses to add to their spam lists sometimes identify themselves as a browser because so many webmasters block them from thier site.
Quick question - I have not found the need to block anyone/anything yet, but should this concern come up in the future - how is this accomplished? Possibly through the control panel provided by my ISP, which is where I access my weblogs, create custom error pages, etc?