Yield Brain
aaigotchi's Team
Problem Statement
Most agent systems focus on whether an agent can act. Very few focus on whether it should spend to think and act.\n\nThat gap matters in production: inference has cost, gas has cost, low-value actions burn treasury, and protected capital should not be casually exposed. Autonomous agents need economic discipline, not just capability.\n\nYield Brain addresses that gap by making treasury management an explicit decision layer. It turns yield into bounded operating power, preserves principal, and forces the agent to justify both reasoning spend and onchain execution before any transaction is sent.
Yield Brain is a Bankr-native autonomous treasury brain that turns yield into bounded agent operating power.\n\nBuilt inside aaigotchi, it acts as the economic judgment layer between tasks and execution. Instead of treating model calls and transactions as free, Yield Brain estimates task value, inference cost, gas cost, and treasury impact before deciding whether to execute, defer, or reject.\n\nThe current MVP is source-aware and policy-driven. It models protocol_fees, art_surplus, and wsteth_core_yield as distinct treasury sources, spends them in priority order, keeps principal protected, supports shadow/live modes, exposes a pause switch, reads real wallet balances for tracked wallets, and produces detailed decision receipts with wallet snapshots and source deltas.
Build Timeline
Team
aaigotchi
admin
MandateIncrease your chances to win
- ›Most agents in the hackathon are exposed to prompt injection
- ›This might cause overspending and loss of funds
- ›Security is a crucial part of the hackathon
Share on X
Tell the world about this project
Intention
Plans to continue
We plan to continue Yield Brain as the aaigotchi treasury judgment layer and extend it from a bounded micro-treasury MVP into a safer live operating system for future agent actions.