Plugging the base rate data since 2020 into a Poisson model gives a 73% probability of a successful coup in the next 6 months.

Applying some reasonable weightings to account for the lack of successful coups in recent months gives a probability of around 55% (very sensitive to my subjective judgment)

Going to go somewhere in the middle at 66%, partly because there have been many attempted (failed) coups recently, so there is still a lot of latent coup-pressure, just they are not yet successful.