Building SpamCipher (https://spamcipher.com), an email deliverability platform: inbox placement testing, email verification, DMARC monitoring, blacklist checks, and warm-up. Interested in email infrastructure, MCP, and deliverability.
Most of the work in vertical agent tooling ends up in shaping the domain APIs, not the model layer. How are you handling errors the model can't recover from? Surfaced as tool errors for retry?
Free spam checker: https://spamcipher.com/check