The question is highlighting the approach of showing that the sequence An converges (to 1) by saying that:
If you give me any e>0, I can find a big number N(e) for that e, such that, after N(e) terms in the sequence, the terms will all be within e from the limit, in this case 1.
That is how convergence of sequences seems to be defined.
So, the sequence is 0, 0.5, 0.66... , 0.75, 0.8, 0.833... etc
If the given e was 0.4, you could let N(0.4) = 2. Then after the 2nd term, all terms are within 0.4 of the limit, which is 1.
If the given e was 0.3, you could use N(0.3) = 3.
So, what you want to do is to find a general expression N(e) that bounds the tail of the sequence to being near enough to 1.
For any n:
|An - 1| = |n/(n+1) - 1|
= 1/(n+1)
If you want this to be less than e, you require that n+1 is greater than 1/e
So, define N(e) = the first natural number greater than 1/e.
This can be called the roof function for 1/e.
Then, we find that, if n is greater or equal to N(e):
=> n > 1/e
=> n + 1 > 1/e
=> 1/(n+1) < e
<=> |An - 1| < e since we earlier showed that |An - 1| = 1/(n+1)
Which is the implication required.