Introducing the Open Knowledge Format: A New Standard for AI Context

Solve the context gap in LLM agentic systems with OKF, a vendor-neutral standard for portable, human- and machine-readable knowledge.

robot, workflow, network

As foundation models continue to improve, they are hitting a critical bottleneck: the lack of relevant context. While modern LLMs can write code or analyze datasets with ease, their ability to function within agentic systems is often limited by the quality of the information provided.

To build truly autonomous agents, we need more than just smarter models; we need a reliable way to deliver curated, actionable knowledge.

The Problem: We Need a Format, Not Another Service

The solution to this context gap isn’t another proprietary knowledge service. When building complex workflows, such as those explored in mastering LLM app development with Dify, you quickly realize that data must be portable and persistent.

You need a way to represent knowledge that satisfies these core requirements:

  • Universal Production: Anyone can create it without needing a specific SDK.
  • Seamless Consumption: Any agent or human can use it without complex integrations.
  • True Portability: It survives moves between different tools, organizations, and systems.
  • Version Control Friendly: It lives alongside your code in git repositories.

Introducing the Open Knowledge Format (OKF)

code, network, folder

Today, we are introducing the Open Knowledge Format (OKF). This is an open specification that formalizes the emerging ‘LLM-wiki’ pattern into a portable, interoperable standard.

The beauty of OKF v0.1 lies in its simplicity. It is vendor-neutral and designed to be both human-friendly and agent-friendly. A bundle of OKF documents is:

  • Just markdown: Readable in any editor and indexable by search tools.
  • Just files: Shippable as a tarball or hostable in any git repo.
  • Just YAML frontmatter: For structured, queryable fields like type, title, and tags.

How OKF Works: The Design

code, network, folder

An OKF bundle is simply a directory of markdown files representing various concepts. Because the file path serves as the concept’s identity, you can create a rich graph of relationships using standard markdown links.

sales/
├── index.md
├── datasets/
│   ├── index.md
│   └── orders_db.md
├── tables/
│   ├── index.md
│   ├── orders.md
│   └── customers.md
└── metrics/
    ├── index.md
    └── weekly_active_users.md

The Anatomy of a Concept Document

Each document utilizes YAML frontmatter for structured metadata and a markdown body for the core content. For instance, you can link directly to a BigQuery table resource within the metadata.

---
type: BigQuery Table
title: Orders
description: One row per completed customer order.
resource: https://console.cloud.google.com/bigquery?p=acme&d=sales&t=orders
tags: [sales, revenue]
timestamp: 2026-05-28T14:30:00Z
---

# Schema

| Column | Type | Description |
|---|---|---|
| order_id | STRING | Unique identifier |

Three Principles Behind the Design

  1. Minimally opinionated: OKF only requires a type field; everything else is up to the producer.
  2. Producer/consumer independence: A bundle hand-authored by a human can be just as easily consumed by an AI agent.
  3. Format, not platform: OKF is never tied to a specific cloud provider or model framework.

The value of a knowledge format comes from how many parties speak it, not from who owns the infrastructure.

Ready to explore the spec? Check out the official OKF specification on GitHub and start building more context-aware systems today.

code, network, folder

Open Knowledge Format (OKF)

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 that knowledge is best represented in commonly accessible, established formats that are:

  • Readable by humans without tooling.
  • Parseable by agents without bespoke SDKs.
  • Diffable in version control.
  • Portable across tools, organizations, and time.

The format is minimally opinionated. It standardizes only the small set of structural conventions needed to make a knowledge corpus self-describing — anything beyond that is left to the producer.

Goals

  1. Define a universal format that enrichment agents can write into.
  2. Inform how consumption agents should read and traverse it.
  3. Facilitate exchange of knowledge across systems and organizations.
  4. Standardize the small number of required fields that must be present for content to be meaningfully consumed.

Non-goals

  • Defining a fixed taxonomy of concept types.
  • Prescribing storage, serving, or query infrastructure.
  • Replacing domain-specific schemas (Avro, Protobuf, OpenAPI, etc.) — OKF references them; it does not subsume them.
Tags