Honeybee pay
AI-assisted crypto payments with buyer-approved rules and zero-knowledge privacy.
Demo
About
Honeybee Pay aims to make everyday crypto payments simple, private, and secure. AI helps users prepare payments in plain language, while a payment firewall enforces the recipient, token, and amount the buyer approved. The AI cannot override those rules. A planned zero-knowledge proof layer will allow payment validity to be verified without publicly revealing selected transaction details. The first demo will compare three payments of 1 test USDC: a normal payment, a redirected payment in an intentionally unprotected demonstration, and the same redirection attempt blocked by the protected system. AI will explain why the payment was blocked. Development is underway; zero-knowledge privacy will be implemented and tested separately from the initial firewall demo.
How it's made
I’m building Honeybee Pay in GitHub Codespaces using Foundry to develop and test Solidity smart contracts. So far, I have documented the payment rules and demo scenarios and installed the development tools. The planned payment contract will check each payment against the buyer’s approval, including the recipient, token, amount, and expiration, and prevent approvals from being reused. AI will help prepare payment requests and explain blocked attempts, but it will not have authority to override the contract’s rules. The initial tests will compare normal payments with deliberately redirected payments using test USDC. The unprotected example will be kept separate from the protected payment contract. Zero-knowledge privacy is a separate planned phase; the proof system and its integration have not yet been selected or implemented. I’m using AI tools for development guidance and coding review. I will document their contributions alongside my implementation decisions, tests, and build progress.