One significant hurdle was developing the trust score mechanism to be unfalsifiable and reliable. This required careful design and multiple iterations to ensure that the system could not be gamed or manipulated by bad actors. I overcame this by implementing a robust feedback loop using non-transferable SFTs, two of which are distributed at the moment of problem resolution or can be transferred by peers to each other by sacrificing their own balance of those. This mechanism accurately reflects users' contributions and interactions.

Another challenge was integrating Farcaster Frames. Ensuring seamless interoperability and data flow between these platforms required extensive testing and adjustments. I resolved this by creating a separate API in my Vercel edge function and employing thorough testing procedures to ensure consistent functionality. The deployment itself was quite a challenge too.

Incorporating Coinbase bundler and paymaster took another noticeable effort because it is yet to properly support ZeroDev Kernel wallets, especially undeployed ones.

Adding mandatory charity contributions into the reward system posed a technical and ethical challenge. Balancing user incentives with charitable giving required fine-tuning the reward distribution algorithm. I addressed this by allowing customizable charity rates within a defined range, ensuring both user satisfaction and meaningful contributions to charity.

Lastly, the biggest challenge was making the submission on time by the scheduled deadline. It was very hard to properly estimate task timing and deliver the best result possible while combining the development with my day job :)

These challenges taught me valuable lessons in problem-solving, iterative development, and maintaining a user-centered focus, ultimately strengthening aHand and motivating me to continue building further. I aim to eventually ship a production version, start growing a community around it, and solve people's problems on a mass scale.