Monsters.fun

Bonding-curve products reduce to buy, watch, sell. That loop keeps active speculators but gives a casual player no reason to open the app when they are not trading — a retention dead end the team wanted to design out.

Prototype a game economy on Abstract where each token is inseparable from an AI character, and where price action is expressed as a consequential game event rather than a chart.

• 01Built the AI Monster behaviour layer on an AI Town-derived agent simulation framework, so each creature has its own decision patterns rather than static stats.
• 02Bound a bonding-curve 'Shard' token 1:1 to each Monster's on-chain identity, progression and stats.
• 03Shipped the Training Dojo: wave-based runs against AI opponents that unlock moves, drive evolution, and measurably refine the AI's own battle decision-making.
• 04Implemented the market-cap transfer mechanic — a winning Monster takes a defined portion of the loser's market cap, making 'my monster won' and 'my token went up' the same event.
• 05Built a weekly points leaderboard paying USDC to the top 50, scoring battles, training consistency, trading, referrals and community activity.

• Playable prototype where token value and game progression are one system.
• Retention loop driven by progression and competition, not price alone.
• Gated Day 0 Access rollout delivered on schedule for launch.
• Incentives spread across five activity types rather than trading volume alone.