avoiding nested window functions?

[NickCrews]

I want to find the minimum row per group, but only if there is a single minimum per group:

import ibis

ibis.options.interactive = True

t = ibis.examples.penguins.fetch()
is_smallest = t.body_mass_g == t.body_mass_g.min()
t2 = t.group_by("island").mutate(
    is_smallest=is_smallest,
    n_smallest=is_smallest.sum(),
)
smallest_per_island = t2.filter(t2.is_smallest, t2.n_smallest == 1)

This gives the duckdb sql

SELECT
  "t0"."species",
  "t0"."island",
  "t0"."bill_length_mm",
  "t0"."bill_depth_mm",
  "t0"."flipper_length_mm",
  "t0"."body_mass_g",
  "t0"."sex",
  "t0"."year",
  "t0"."body_mass_g" = MIN("t0"."body_mass_g") OVER (PARTITION BY "t0"."island" ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING) AS "is_smallest",
  SUM(
    CAST("t0"."body_mass_g" = MIN("t0"."body_mass_g") OVER (PARTITION BY "t0"."island" ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING) AS INT)
  ) OVER (PARTITION BY "t0"."island" ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING) AS "n_smallest"
FROM "penguins" AS "t0"

which results in Binder Error: window function calls cannot be nested

1. is there some SQL that can actually do what I want in one go? I can get the result I want by chaining two subsequent group_by()s, finding the min values the first time, then counting them the second time, but I think if I did them at the same time that would be faster. eg
2. if that SQL exists, then could we make it so Ibis generates it?

[cpcloud]

There's no SQL that can do that in a single query of which I am aware.