Conviction

Conviction — a narrative intelligence terminal for crypto with an onchain conviction record. It pulls ~150 live trending tokens from DexScreener, groups them into narratives (AI, RWA, DeFi, Memecoins, Gaming, DePIN), and scores each one 0–100 based on liquidity, volume, momentum, and breadth. The scoring is deliberately transparent. Then the onchain part: you can commit your own call on any narrative (bullish/bearish + a 280-char thesis) to a ConvictionRegistry smart contract I wrote and deployed on Monad testnet. Every call is immutable and timestamped, displayed in a live feed read straight from the chain. The problem (a personal one): I've always wanted to be early to a narrative that's heating up onchain and socials (Robinchain as a case study) and I don't have much time keeping up with social because of my web2 job and honestly this still a working progress and I plan on making it better. What I learned: This was my second smart contract deployment, I learnt writing, deploying, and integrating a Solidity contract again (via viem) in one weekend. I also learned that live market data fights back: I debugged a broken data pipeline, deduped tokens, and had to filter out a troll token whose name was 1,000 characters of concatenated tickers that exploded my layout. Real data is messy; shipping means handling it.