COVID-19 pandemic, starting from the latest 2019, and caused by SARS-CoV-2 pathogen, led to the hardest health-socio-economic disaster in the last century. Despite the tremendous scientific efforts, mainly focused on the development of vaccines, identification of potent and efficient anti-SARS-CoV-2 therapeutics still represents an unmet need. Remdesivir, an anti-Ebola drug selected from a repurposing campaign, is the only drug approved, so far, for the treatment of the infection. Nevertheless, WHO in later 2020 has recommended against its use in COVID-19. In the present paper, we describe a step-by-step in silico design of a small library of compounds as main protease (M
pro
) inhibitors. All the molecules were screened by an enzymatic assay on M
pro
and, then, cellular activity was evaluated using Vero cells viral infection model. The cellular screening disclosed compounds
29
and
34
as in-vitro SARS-CoV-2 replication inhibitors at non-toxic concentrations (0.32 < EC
50
< 5.98 μM). To rationalize these results, additional in-vitro assays were performed, focusing on papain like protease (PL
pro
) and spike protein (SP) as potential targets for the synthesized molecules. This pharmacological workflow allowed the identification of compound
29
, as a dual acting SARS-CoV-2 proteases inhibitor featuring micromolar inhibitory potency versus M
pro
(IC
50
= 1.72 μM) and submicromolar potency versus PL
pro
(IC
50
= 0.67 μM), and of compound
34
as a selective SP inhibitor (IC
50
= 3.26 μM).