@modelguardnet

ModelGuard is building an insurance layer for AI agents in crypto, focusing on verifying agent intent, behavior monitoring, and underwriting coverage.

They've partnered with Native Insurance (a Lloyd's-connected broker) to offer actual insurance products, and are developing standardized "Intent Reports" based on AIUC standards.

The project has a $GUARD token, is live on ACP Sandbox with their Genesis Ratings agent for testing, and appears to be in early mainnet deployment phase with real product iteration.

Green flags: Partnership with established insurance broker (Native Insurance / Lloyd's connection) for real coverage · Working product on ACP Sandbox with Genesis Ratings agent deployed · Technical substance: building on AIUC standards, publishing detailed intent reports for projects like Wach AI · Small following (518) with recent account creation (June 2025) indicating genuine early-stage · Clear differentiation: first-mover in AI agent insurance for crypto-native systems

Red flags: Very low engagement (avg 23) relative to ambitious claims about insurance underwriting · No accessible website data to verify claims independently · Token ticker mentioned but unclear if live/tradeable or just announced · Heavy reliance on partnerships and other projects for credibility rather than standalone traction