Selecting one of multiple values that satisfy a condition

So I have this data frame of a condition column and a value column. Suppose I want to change one of the values for which the condition is satisfied, but not all of the values for which the condition is satisfied. The idea here is that I would like to change the values one by one (or select one randomly).

df <- data.frame(condition = c(TRUE,TRUE,TRUE,TRUE,FALSE,FALSE,FALSE,FALSE),
                 value = 8:1)

I would like to store this in a new column. Suppose I randomly select row 2: The condition is TRUE and the original value is 7. I change it to 0 (or some other value). The desired output in this case is:

> df
condition value
1      TRUE     8
2      TRUE     7
3      TRUE     6
4      TRUE     5
5     FALSE     4
6     FALSE     3
7     FALSE     2
8     FALSE     1

I suppose I could do this with an if statement, a third column of group-row-numbers and the sample function, but is there a nicer way? In particular, I need to make sure I do not generate a new column in case the condition is not satisfied for any of the values. Otherwise I am wasting a lot of computing time later on in the analysis. I usually do things inside the tidyverse if I can, but I do not mind if there is a nice solution without using the package (or using a different package). Thanks!