Ok, I know everyone and their moms is going to be posting on this but, as many of you have noted, I'm heavily biased towards maniraptorans and this is one of the coolest of the decade.
Hyperbolic Jurassic Park fanboys, meet Balaur bondoc, the dromaeosaurid with not one, but two sickle claws on each foot.
Above: Boom. From NatGeo.
Note that the two sickle claws are on digits one and two, and digit one is essentially anti-retroverted, pointing forwards as in therizinosaurs. Balaur also has a lot of fusion in the hand, with digit two and three fused together and digit 3 reduced, as in caudipterids. Balaur lived on the island of Hateg in Maastrichtian (latest Cretaceous) Transylvania, a location known for its insular dwarfism among herbivorous dinosaurs, including the dwarf sauropod Magyarosaurus and dwarf hadrosaur Telmatosaurus. While Balaur is about the size of the larger contemporary dromaeosaurs at around 2 meters in length, its unique suite of derived skeletal characters also fits into the "island rule," according to Sues in an accompanying write-up to the official paper. While herbivorous forms tend to "shrink" on islands to conserve resources, predators often grow larger to better exploit the dwarf herbivores, relatives of which would be out of their league elsewhere. A good example of this are the famously small, extinct Stegodon dwarf elephants of Flores (or indeed the apparently dwarf humans, Homo floresiensis), and the contemporary giant monitor lizards like the Komodo dragon. However, no teeth of any carnivore larger than Balaur have been found in the Hateg basin deposits. Teeth are usually the most numerous and obvious indicators of the local carnivore population, so Balaur was probably the largest predator in its ecosystem. This would seem to fit with its stocky build and double sickle claw: here was an animal that truly met the popular image of dromaeosaurs grappling with prey larger than themselves. The extra claws and solid build of Balaur would have come in handy when taking down a hadrosaur or sauropod.
You can read more on this find at National Geographic.
Update: It seems like there's some confusion about the name. Tom Holtz on the DML, and some parts of the above NatGeo article, reported it as Baldaur bondac. Others (and other parts of the linked article) use Balaur bondoc. The later is he correct spelling.