In this post, there is a counterexample where a nil ideal (ie every element is nilpotent) is not nilpotent.
More precisely, let's consider the polynomial ring $$R:=\Bbb C[X_1,X_2,X_3,\dots]/\langle X_1^2,X_2^3,X_3^4,\dots\rangle,$$ and set $I:=\langle X_1^2,X_2^3,X_3^4,\dots\rangle$. Then we can consider the ideal $$\tilde I:=\langle X_1+I,X_2+I,X_3+I,\dots\rangle \trianglelefteq R.$$ But I can not see why this is a nil ideal. Normally, we should take an arbitrary element $\bar x$ of $\tilde I$ and show that there exists an $m\in \Bbb Z^+$ such that $\bar x^m=0_R=I$.
Could you please help?
Thanks.
It seems that your definition of a nil ideal is that every element is nilpotent.
It is an easy exercise that the sum of two nilpotent elements is still nilpotent.
Solution:
Using induction, one can deduce that any finite sum of nilpotent elements is nilpotent.
Equally clear is that any multiple of a nilpotent element is also nilpotent.
Therefore, any ideal generated by (possibly infinitely many) nilpotent elements is nilpotent, because every element in the ideal is a finite sum of multiples of the generators.
This in particular applies to your $\tilde I$, which is generated by images of the $X_i$'s.
It seems that your definition of a nilpotent ideal is that there is an $n$ such that $I^n = 0$.
Then it's clear that $\tilde I$ is not nilpotent, since for every $n$, the ideal $\tilde I^n$ contains the image of $X_n^n$, which is nonzero in $R$.