question: how can I rate limit the use of edge functions? e.g. let say some malicious actor finds the endpoint client-side and spam it, how can I prevent it to avoid wasting edge function calls?