Passage Protocol

A cryptographic logging system that creates verifiable records when AI agents enter or leave contexts, signed with Ed25519 or P-256 keys. Think of it as an audit trail for agent sessions where each departure and admission gets a cryptographic signature you can verify later. You'd reach for this when you need tamper-proof logs of which agents accessed what systems and when, especially in multi-agent scenarios where accountability matters. The signature schemes are standard enough that other systems can verify these records independently. Useful for compliance logging, debugging agent behavior across sessions, or building trust boundaries between different AI systems that need to prove their interaction history.