I am concerned about the security risks involved in using your service. I would appreciate it if you could consider implementing some of the following suggestions:
- Use a more specific user agent than "go-http-client" to identify your requests.
- Provide a list of verified IP addresses that you use to send webhooks, so that I can restrict access to my server with a firewall.
- Use a dedicated IP address or a domain name to send webhooks, instead of a shared one.
- Then you might even verify yourself to cloudflare (to be recognized as known bot)
These measures would greatly enhance the trust and reliability of your service, and I hope you can take them into account for future improvements.
I understand that you may have some limitations due to your architecture design and security policies, but I hope you can at least acknowledge the importance of these issues and provide some feedback on the feasibility of these suggestions.
I think these suggestions do not actually improve security. The only way to make sure that content can be trusted is through authentication or by signing the message.
For instance: A client can easily fake any user agent, so unless the user agent contains a secret that is only shared with you and Mailgun, this does not improve security at all.
Or an ip address can easily be spoofed, so I can easily get around that firewall.