Why does pool function in r give "Error: $ operator is invalid for atomic vectors" error?

My first question on Stackoverflow...sorry for any inconvenience and imprecision. I'm performing an analysis on multiple imputated dataset, generated with mice package. I used split function to generate the data.frame list and, on each data.frame, I added an IPTW variable estimated on each data.frame separately. With my data ready, I performed the models of interest (weighted Cox models with cph function) and put the results in a list (fit0). When I try to run

pool(fit0)

I get the message:

Error: $ operator is invalid for atomic vectors

I looked into pool function and I find the problem could be in pool.fitlist subfunction. At the beginning of the pool.fitlist, there's a summary that only want extract the coefficients of the models from my list, but if I try to run it out of the function (after a as.mira transformation of my list of results) I get again the error message:

Error: $ operator is invalid for atomic vectors

Looking to my results, I found that

> is.atomic(fit0$analyses[[2]]$coefficients)
[1] TRUE

Any idea how to solve this little drawback?

EDIT: I add the code and the fake data of example. Initially I create 4 short data.frame and put them in a list

n <- 6
d1 <- data.frame(id=1:n, 
                 event=c(0,1,1,1,1,1),
                 time=c(19.67, 11.94, 7.86, 1.58, 48.55, 36.18),
                 age=c(59,75,38,67,68,41),
                 drug=c("Yes","No","Yes","Yes","Yes","Yes"),
                 w=c(0.07229, 0.52002, 0.04708, 0.06576, 0.08355, 0.04033)
)
d2 <- data.frame(id=1:n, 
                 event=c(0,1,1,1,1,1),
                 time=c(19.67, 11.94, 7.86, 1.58, 48.55, 36.18),
                 age=c(59,75,38,67,68,41),
                 drug=c("Yes","No","Yes","Yes","Yes","Yes"),
                 w=c(0.0687, 0.5395, 0.0289, 0.1070, 0.1536, 0.1178)
)

d3 <- data.frame(id=1:n, 
                 event=c(0,1,1,1,1,1),
                 time=c(19.67, 11.94, 7.86, 1.58, 48.55, 36.18),
                 age=c(59,75,38,67,68,41),
                 drug=c("Yes","No","Yes","Yes","Yes","Yes"),
                 w=c(0.01654, 0.44411, 0.06588, 0.07772, 0.05326, 0.03234)
)
d4 <- data.frame(id=1:n, 
                 event=c(0,1,1,1,1,1),
                 time=c(19.67, 11.94, 7.86, 1.58, 48.55, 36.18),
                 age=c(59,75,38,67,68,41),
                 drug=c("Yes","No","Yes","Yes","Yes","Yes"),
                 w=c(0.02490, 1.00000, 0.21502, 0.08058, 0.11399, 0.07635)
)

ld <- list(d1,d2,d3,d4)

After than I generate preform the IPTW models and gather the results in a list and use the pool function:

modelFit0 <- vector("list", 4)

for(i in 1:4) {
  
  res <- cph(Surv(time,event)~ rcs(age,3) + drug,
              data=ld[[i]], weights=w)
  modelFit0[[i]] <-res

}  

pool.fit.OS <- pool(modelFit0)

Here is where I get:

> pool.fit.OS <- pool(modelFit0)
Error: $ operator is invalid for atomic vectors
In addition: Warning message:
In get.dfcom(object, dfcom) : Infinite sample size assumed.

Hope this is a valide reproducible example.

I use cph packeage from rms library. These are my sessionInfo:
R version 4.1.2 (2021-11-01)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19043)
other attached packages: rms_6.2-0 [...]