This is a front-end for the Online Encyclopedia of Integer Sequences, made by Christian Perfect. The idea is to provide OEIS entries in non-ancient HTML, and then to think about how they're presented visually. The source code is on GitHub.
%I A062802 #25 Oct 19 2017 10:46:14
%S A062802 2,11,29,2999
%N A062802 a(1) = 2; a(n+1) = smallest prime > a(n) whose sum of digits is a(n).
%C A062802 a(5) = 5*10^333 - 10^332 - 10^174 - 1 =
%C A062802 489999999999999999999999999999999999999999999999999999999999999999999999999\
%C A062802 99999999999999999999999999999999999999999999999999999999999999999999999\
%C A062802 99999999999998999999999999999999999999999999999999999999999999999999999\
%C A062802 99999999999999999999999999999999999999999999999999999999999999999999999\
%C A062802 9999999999999999999999999999999999999999999999
%C A062802 which is too large to include in the DATA field. - _Don Reble_, Sep 03 2006
%C A062802 Define b(n), n>=0, to be the smallest prime p such that applying the sum-of-digits function n successive times to p produces n distinct primes (excluding p itself). Is b(n) = a(n) for all n? The first four terms agree. - _Felix Fröhlich_, Aug 13 2015
%C A062802 It is very likely that this is the case, since although there are always larger "parent" primes with the same digital sum, they typically are at least twice as large (for p=2, these are 11, 101, ...; for p=11 these are 29, 47, 83, ...; for p=29 these are 2999, 3989, 4799, ...), and the number of *digits* of the next term is roughly proportional to this value, so even the "second best" choice would typically lead to a much larger "parent" prime. - _M. F. Hasler_, Aug 16 2015
%t A062802 a = {2}; k = 3; Do[While[Total@ IntegerDigits@ k != a[[n - 1]], k = NextPrime@ k]; AppendTo[a, k], {n, 2, 4}]; a (* _Michael De Vlieger_, Aug 20 2015 *)
%Y A062802 Different from A103830 after a(4).
%K A062802 nonn,base
%O A062802 1,1
%A A062802 _G. L. Honaker, Jr._, Jul 19 2001