Security-First: Our 4-Layer Risk Architecture

In DeFi, security is not a feature -- it is the foundation. Every protocol that has lost user funds did so because security was treated as an afterthought. At 963X, the risk architecture was designed before the first line of execution code was written.

Layer 1 -- Static Rules: Every transaction is checked against hardcoded constraints: gas price limits, rate limiting per wallet, maximum transaction size, and failure thresholds. These rules cannot be overridden by governance or agents.

Layer 2 -- Dynamic Risk: Volatility-based adjustments reduce position limits during high-volatility periods. If BTC moves 5% in an hour, maximum leverage is automatically reduced across all BTC markets.

Layer 3 -- Behavioral Analysis: Statistical anomaly detection flags unusual patterns -- wallet activity that deviates more than 3 standard deviations from historical norms triggers additional verification.

Layer 4 -- Circuit Breakers: System-wide emergency mechanisms that pause all execution if aggregate risk metrics exceed thresholds. This is the last line of defense against systemic risk events.

AI agents operate under additional security constraints. They never hold private keys. They submit transaction requests through the same 4-layer pipeline. Their blast radius is capped by design -- a compromised agent can affect at most 0-5% of user portfolio value.

Permission grants follow the principle of least privilege. Users specify exactly what actions an agent can take, with what assets, up to what size, and for how long. Permissions can be revoked instantly.