More radare 2 (5/?): environment variables etc.

I'm still focusing on the narnia
(https://overthewire.org/wargames/narnia/) capture the flag,
which is trivial in itself, but I'm trying to beat it completely
within R2, with the goal of learning the tool.

Radare2 is rich of features, but also baroque and inconsistently
documented.  How do I set environment variables?  Rarun is the
answer, but the documentation glosses over how that is
integrated.

Search on the web.
Find out about radare2-explorations (a complementary book)
https://monosource.gitbooks.io/radare2-explorations/content/intro/debugging.html

In a nutshell:

1. Run `rarun2 -t` in another tty and yank the path of the
   printed character device (e.g. /dev/pts/8)

2. Execute r2 like this:

   $ r2 -r rarun.rr2 -d program

Example with narnia1:

   $ cat rarun.rr2
   stdio=/dev/pts/8
   setenv=EGG=exploit_here

   $ r2 -r rarun.rr2 -d ./narnia1

Well, the software is powerful and intriguing, but I'm a little
frustrated by the (well known) steep learning curve, IMO mostly
due to old/inconsistent docs and opcode-like commands. For a
person like myself, who has so little free time, this is a big
obstacle.  I'm grateful nonetheless.

--
mov esi, str._bin_sh
mov edi, str._bin_sh
mov eax, 0
call sym.imp.execl