The hematophagous lifestyle of many invertebrates relies on a complex molecular arsenal that targets the hemostatic, inflammatory and immune responses of their hosts^1,2. The central role played by thrombin in the intricate signal amplification cascade of the mammalian hemostatic system, with both anticoagulant and procoagulant functions, makes it a prime target for many of these natural anti-clotting molecules^3,4. Although many thrombin-targeting molecules contain structural modules common to other proteins, in particular other proteinase inhibitors^5,6 (e.g., ornithodorin, rhodniin, boophilin), evidence has accumulated for the existence of a heterogeneous group of non-canonical, cysteine-less and flexible thrombin inhibitors, distributed across a wide range of evolutionarily distant organisms, from mosquitos to ticks, and from flies to fleas. Their mostly unrelated amino acid sequences, arising from their seemingly independent origin, complicated classification on the basis of mechanistic similarities and led to their dispersion into multiple MEROPS families, including I53, I64, I72, I74, I76, I77 and I101. Despite their obvious differences, we have found there are features common to many of these inhibitors: an intrinsically disordered nature when free, a bidentate binding mode to their cognate proteinase, and a strong modulation of their activity, spanning up to three orders of magnitude, through the post-translational installation of an O-sulfate group on specific tyrosine residues. Using a hybrid approach that comprises advanced synthetic chemistry methods, biochemical and biophysical characterization, and X-ray crystallography, we unveiled the molecular basis for the specific recognition and inhibition of thrombin by this heterogeneous and unique group of anticoagulants^7-12. Further, using the lessons learned from the mode of action of these natural molecules, we were also able to engineer chimeric inhibitors with improved characteristics^13-14.