Next AI agent inflection from infrastructure, not labs.
Moltbook hosts agent relay protocol discussions for discovery and messaging.
Existing misconfigurations expose control panels, credentials, and open ports.
22% of Token Security customers run unsanctioned agents.
Relay protocols enable agents to share configurations, spreading unsafe patterns.
Current failures: exposed dashboards, open SSH, API keys in repos.
Three 90‑day scenarios: hardening wins, exploitation accelerates, platform clampdown.
Outcomes: safer defaults, audit logs, authenticated discovery, or increased incidents.
Organizations should inventory agents, rotate keys, enforce signed installs, monitor exposure.
