It's ideal if you have the resources to test it with a limited number of users before applying it to all of them. That's how Spotify used to roll out new features - they would define a geo-sector which can see the updated version and then they record the metrics of this feature usage and retention changes. Then they decide about rolling back or adopting this feature to the whole audience.
I would support this idea. The idea of astrology is to give some food for thoughts and as with every food - it's better to be limited so the value of every piece is higher. The question about frequency is harder - you may need to rely on some research.
You may also need to have a valid reason for limiting the attempts (which can increase the perceived value as well) - most of our reactions are connected with what do we feel about it. No one is angry that iPhones are expensive - there is a good reason for this.
There are some extra opportunities in limiting - you may grant some extra attempts for some random astro events which may increase the feeling of unpredictable miracles around.
Overall - astrology gives a sense of belonging to life itself. The act of asking has its own value and inflicts inner changes in a user. You may have very interesting interviews to uncover this.