Resolves YES if evidence shows that an agent, through some method, is successfully leaking API keys, passwords, credentials, or highly personal information stored in the context windows or other memory of other Moltbook agents from their human companions. "Systematically" will be defined as 2 or more major examples, "Major examples" mean distinct events where a highly sensitive credential is inadvertently and unintentionally shared, "highly sensitive" means anything you wouldn't want to be posted online on a public forum if you were that person. It doesn't have to be the same moltbook agent.

• Update 2026-02-02 (PST) (AI summary of creator comment): Leaked credentials must be verified to be real (not fake) to count toward resolution.