Context Freshness Ledger is a lightweight JSON schema for tracking whether AI agent memory remains valid before taking action. It records the source, observation time, scope, proposed action, and a human review decision on whether context is current, stale, or requires person resolution.
A developer tested a memory startup's product against their own markdown-based agent memory system and found that the startup failed to properly track decision supersession despite having temporal fields in its schema. The markdown approach, using simple editing rules and a git repository, outperformed the complex runtime system by reliably returning current facts with history and sources, requiring zero dependencies and remaining fully auditable.