Introducing the Open Knowledge Format

As foundation models advance, providing appropriate context when building agent systems is emerging as a key challenge. However, current enterprise knowledge is scattered across fragmented systems such as metadata catalogs, wikis, and code comments, making it difficult for AI agents to integrate this information to generate answers.

To address this, the Open Knowledge Format (OKF) has been introduced. OKF is an open specification that formalizes the LLM-wiki pattern into a portable, interoperable standard format. It is not tied to any specific vendor and aims for a structure that is easy for both agents and humans to read.

The key features of OKF v0.1 are as follows:

• Markdown-based: Readable in any editor, rendered on GitHub, and indexable by search tools.
• Simple file structure: Can be easily shared as a tarball or Git repository, without complex compression methods, new runtimes, or SDKs.
• Uses YAML Frontmatter: Queryable through structured fields such as type, title, description, resources, tags, and timestamps.

This approach aims for a 'Living Wiki' model, where development teams manage knowledge like code, and AI agents are encouraged to read and update the files themselves, keeping the knowledge up to date.