NoGhostDrop

Demo

About

The problem it solves Airdrops have become a standard method for user acquisition, especially in new chain launches and protocol marketing campaigns. However, they are increasingly being exploited by bots and sybil wallets, often resulting in token farming without real user engagement. This not only dilutes the value of the token but also reduces the effectiveness of user acquisition efforts — fundamentally defeating the original purpose of airdrops: onboarding real users. NoGhostDrop addresses this by using on-chain behavior analysis powered by LLMs to distinguish between legitimate and sybil wallets. Projects or foundations can create an airdrop, define custom eligibility conditions, and let our system handle everything from wallet evaluation to token distribution. All users need to do is connect their wallet and click “Claim” — the AI handles the rest. This allows: * Foundations to increase the ROI of their marketing campaigns. * Projects to ensure their tokens are going to real, engaged users. * Developers to scale trustless airdrops across chains with minimum friction. By combining off-chain intelligence with on-chain execution, NoGhostDrop offers a secure, scalable, and user-friendly way to run smarter airdrops.

How it's made

Challenges we ran into One of our biggest challenges was integrating ElizaOS and effectively prompting the LLM to evaluate wallet behaviors. Since most of us were unfamiliar with the tool, three of our team members dedicated two full days to learning its structure, setting up the environment, and iterating on prompt engineering. It took multiple trials to craft prompts that could accurately classify wallets based on on-chain activity and airdrop conditions. Another challenge was fetching on-chain data quickly and reliably. The amount of historical and interaction data needed for analysis was substantial, and a single-threaded approach wasn’t cutting it. To improve performance, we implemented parallel RPC requests and optimized data filtering to reduce unnecessary calls — making the wallet evaluation noticeably faster. We also had to carefully design our smart contract architecture to prevent abuse. We implemented a relayer-based execution model, token locking, restricted transfer logic, and creator-only withdrawal to ensure airdrop funds could not be exploited — even if contract calls were triggered externally. This required close coordination between backend logic and Solidity development to maintain security without compromising automation. Despite these hurdles, we iterated quickly, split responsibilities based on our strengths, and built a working demo that not only detects sybil wallets but also distributes tokens in a fair, one-click flow.