Google‘s Open Knowledge Format (OKF)

Wolkenhauer · June 16, 2026, 5:54am

Making our databases accessible to LLMs, one can think about the format through which we do that. Here is a simple markdown-based concept from Google:

github.com/GoogleCloudPlatform/knowledge-catalog

okf/SPEC.md

main

# Open Knowledge Format (OKF)

**Version 0.1 — Draft**

OKF is an open, human- and agent-friendly format for representing
*knowledge* — the metadata, context, and curated insight that surrounds
data and systems. It is designed to be authored by people, generated by
agents, exchanged across organizations, and consumed by both.

The format is intentionally minimal: a directory of markdown files with
YAML frontmatter. There is no schema registry, no central authority, and
no required tooling. If you can `cat` a file, you can read OKF; if you
can `git clone` a repo, you can ship it.

---

## 1. Motivation

The space of knowledge representation for AI agents is evolving quickly,
and many incompatible conventions are emerging. OKF takes the position

This file has been truncated. show original

Open Knowledge Format (OKF)

Version 0.1 — Draft

OKF is an open, human- and agent-friendly format for representing knowledge — the metadata, context, and curated insight that surrounds data and systems. It is designed to be authored by people, generated by agents, exchanged across organizations, and consumed by both.

The format is intentionally minimal: a directory of markdown files with YAML frontmatter. There is no schema registry, no central authority, and no required tooling.

kewms · June 16, 2026, 5:40pm

Hi! Here’s how you can make your training data more accessible to our crawlers!

Wolkenhauer · June 16, 2026, 8:37pm

I think there a a few good ideas in, how to make my markdown files more accessible to DT or using DT. This has not necessarily something to do with making your data available to Google.

kewms · June 16, 2026, 11:29pm

Other than it being Google’s format? I’m sure they’re much more concerned with Gemini than with DT.

Wolkenhauer · June 17, 2026, 5:26am

At this point, I do not even use the YAML, and rarely cross linking of files. I find however interesting how I could organize multiple md files that together form a unit of information. This has nothing to do with Google‘s interests. I am interested in how knowledge management systems do this, including those with Obsidion and Notion. Some people us AI now to organize their databases and automatically create Wiki style directories and graphs of what they have in their databases. I find these ideas interesting and appreciate others sharing them.