Error: `n()` must only be used inside dplyr verbs, not even using n() in my code

I got this message error "Error: n() must only be used inside dplyr verbs" from my code. i've already check answers of this bug on internet, but I can't apply any answers to my situation.

Here is the initial code

#create a fubction to get df with only absorption values, in wide, each column for each wl value
library(tidyverse)
transform_wide <- function(data, value, name){
  df <- data %>% group_by(lambda) %>% 
    mutate(row = row_number()) %>% 
    pivot_wider(names_from = lambda, values_from = {{value}}) %>% 
    select(-row, - campagne, - depth)
  colnames(df) <- gsub(' ', '',paste(name, colnames(df), sep = ''))
  return(df)
}

#apply it on each type of abs
real_df <- transform_wide(real_df, real, 'real')

and the dataset looks like this

real_df <- tibble('campagne'= rep('test', 10),
                  'depth'= c(1:10),
                  'lambda' = c(rep(440,5), rep(470, 5)),
                  'real' = runif(10))

It worked few months ago, so I think it might be due to some dplyr version.

Thanks for your help.