Sortase A (SrtA), a cysteine transpeptidase critical for surface protein anchoring in Gram-positive pathogens, represents an attractive antivirulence target. While covalent SrtA inhibitors show therapeutic potential, existing compounds lack species selectivity. Through structure-guided design, we developed T10, a covalent inhibitor selectively targeting Streptococcus pyogenes SrtA (SpSrtA) over Staphylococcus aureus SrtA (SaSrtA). Molecular docking revealed that shortening a "C=C" bond in lead compound ML346 eliminated SaSrtA inhibition due to steric hindrance from W194, while maintaining SpSrtA binding. X-ray crystallography confirmed T10's covalent modification of Cys208 in SpSrtA. T10 demonstrated two fold enhanced inhibitory potency and species-specific disruption of M-protein anchoring and biofilm formation in Streptococcus pyogenes, without affecting Staphylococcus aureus viability. In a Galleria mellonella infection model, T10 conferred potent protection against lethal infection. This work demonstrates the development of narrow-spectrum antivirulence agents through a structure-based rational strategy.