Show HN: Vera – Open-Source Zero Trust Protocol for AI Agents (12 Services) (berlinailabs.de)

A new zero trust protocol for AI agents, called VERA (Verified Execution & Reputation Agent), has been announced following an upgrade from the previous Agent Trust Framework (ATF). This shift arises from the emergence of challenges in centralized trust mechanisms, highlighted by the collapse of Moltbook's social layer, which underscored the need for sovereign and decentralized trust. VERA embeds trust directly into the agent's execution layer through cryptographic Proof of Execution (PoE), eliminating potential single points of failure and enhancing security. The VERA protocol not only introduces advanced features like support for Ed25519 and post-quantum ML-DSA-65 signatures, ensuring longevity and validity against future quantum threats, but it also includes a Reputation Bridge that automatically boosts an agent's reputation based on verified actions in a sybil-resistant manner. The entire 12-service suite has been refactored to meet VERA's stringent architecture, emphasizing a robust Sovereign Node responsible for key management and peer-to-peer broadcasting. The open-source nature of the VERA implementation allows developers to deploy their own nodes, fostering community involvement and innovation in the AI/ML landscape.