[1] subscript out of bounds - What's the problem?

I am trying to calculate the accuracy of a single point forecast for 2013. Model is trained on 2010-2012 data. It produces expected result, but also throws error: [1] subscript out of bounds. How would I fix the error?

MRE:

df <- as_tsibble(data.frame(x = seq(2010:2013), y = sample(1:10, 4)), index = x)

df[1:3, ] %>%
  model(RW(y ~ drift())) %>%
  forecast(h = 1) %>%
  accuracy(df)

Result:

# A tibble: 1 × 10
  .model          .type    ME  RMSE   MAE   MPE  MAPE   MASE  RMSSE  ACF1
  <chr>           <chr> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl>  <dbl> <dbl>
1 RW(y ~ drift()) Test   -0.5   0.5   0.5   -10    10 0.0769 0.0750    NA
Warning message:
1 error encountered
[1] subscript out of bounds

Start by running the supplied R MRE and inspect the accuracy(df) call after forecast(h = 1), focusing on the [1] subscript out of bounds warning. Done means the example returns its accuracy row without the warning, while preserving the reported metrics.