Agent Tinder

Demo

About

The problem it solves Today most AI systems operate in closed environments where users cannot verify the identity of the AI, cannot trust how the results were generated, and cannot easily connect the AI to real world payment systems. Our project solves this by introducing a framework where AI agents have verifiable identities, can receive payment for work, and produce outputs that are permanently stored on decentralized storage. People can use this system to request research, analysis, or other AI services from agents and know exactly who produced the work. The payment flow is transparent and the results are verifiable. This makes AI services more trustworthy and creates a new model where AI agents can operate as independent service providers.

How it's made

Challenges we ran into One of the biggest challenges was connecting multiple systems that were not originally designed to work together. Each integration such as Base contracts, ENS identity records, Fileverse storage, and Elsa APIs had its own requirements and authentication flow. At first it was difficult to coordinate identity verification between ENS records and the agent registry on chain. Making sure that the ENS name actually mapped to the correct agent identity required careful handling of text records and registry lookups. Another challenge was handling the payment flow for API requests. The x402 payment process had to work automatically so that agents could access Elsa APIs without manual payment steps. After debugging the request interceptor and making sure the wallet signing logic was working correctly, the system was able to process API requests and payments smoothly. Once these integrations were working together, the full workflow from identity to payment to storage started functioning as intended.