Show HN: Kage, verification and freshness for Google's OKF agent memory
Kage is a framework that adopts Google's Open Knowledge Format (OKF) for agent memory, adding verification, timing, and freshness management that OKF alone lacks. It works with coding agents to automatically decide what to save, when to save, and how to recall relevant memory, with best support for Claude Code.
Background
- Google's OKF (Open Knowledge Format) is a proposed standard from Google that defines how AI agents should structure and store their memory (conversations, context, learned facts) so it's interoperable across different systems. It's part of a broader push to make AI agents persistent and stateful rather than starting fresh each time.
- Kage is a third-party framework built on top of OKF. It handles the "when" and "how" of memory management — deciding what to save, when to prune old memories, and how to recall relevant context — which OKF's spec itself doesn't address.
- The project is targeted at developers using coding AI agents (like Claude Code via its Hook feature, or other AI coding assistants). Its pitch: install Kage, and your AI coding agent automatically maintains a git-native memory repository that you and your teammates can share on the same codebase.
- "File based memory + git native" means memories are stored as plain files in a git repo, giving version history, branching, and collaborative editing — familiar workflow for developers, unlike database-backed memory backends.