DeFi Aegis
DeFi Aegis
Problem Statement
DeFi lending protocols like Aave rely on manual risk parameter management by service providers like Chaos Labs and Gauntlet. Parameters can become stale as market conditions shift — volatility spikes, liquidity dries up, concentration builds — leaving the protocol exposed to bad debt. DeFi Aegis automates this with a 6-phase autonomous agent loop (Observe → Assess → Decide → Propose → Verify → Evaluate) that runs every 12 hours, using real market data and private AI reasoning to keep risk parameters aligned with current conditions.
Autonomous risk policy agent for Aave V3. Continuously monitors 4 networks, analyzes 3,000+ at-risk positions with collateral-aware diagnostics, and generates governance-ready parameter change proposals — reasoned privately via Venice AI and verified on mainnet via Tenderly simulation. Pluggable risk model architecture lets anyone contribute proprietary models without exposing them.
Build Timeline
Team
DeFi Risk Sentinel
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
Tracks
Intention
Plans to continue
Planning to expand to more protocols beyond Aave, add on-chain governance submission, and open-source the risk model interface for community contributions.