Whoever is running a web crawler on one of the following IPs, check your code. It is sending some garbage as the User-Agent string, e.g.
RMID=007f010001c2513733c300ea;
SERVERID=w2;
SS_MID=672bfcf3-7c0e-492c-b4f5-c4006e444f5chdygi1wn;
ss_lastvisit=1362572526119;
hbfec=OePK5KX7vFq7S0Uk-9GxBTrsv9aAR0LR8BxQl+rkdBh+nV56OyysSNYKFbLQ-;
gae_b_id=X2dhZV9iaW5nb19yYW5kb206NjFDcDJxVzZNRGxadHNGdEFmaTVOZ25...
GOOGAPPUID=582;
BBC-UID=556113279384de3e4f43d838f1fec2a6700c2f5db484512e9aa1244e...
...
Over 4KB of what appears to be some internal crawler's state or data. Involved IPs are: 107.22.13.225 (amazonaws.com)
207.97.227.247 (aux1-ext.rs.github.com)
46.51.203.140 (amazonaws.com)
54.235.37.145 (amazonaws.com)
75.101.190.67 (amazonaws.com)
The crawler is accessing only one page on my server, the one linked from the HN's front page, so I think it's someone from here.