Open Knowledge Format: A Portable Standard for AI Context Sharing

The article exposes the fragmentation of context that limits AI agents in organizations. Internal knowledge—table schemas, metric definitions, runbooks—lives in metadata catalogs, wikis, code comments, and senior engineers’ heads, each with incompatible APIs and schemas. Every agent builder must solve the same context-assembly problem from scratch, and knowledge remains locked in silos, preventing portability across teams and tools.

The concrete technical path is the Open Knowledge Format (OKF), an open specification that formalizes the emerging LLM-wiki pattern into a portable, interoperable format. OKF represents knowledge as a directory of markdown files with YAML frontmatter, encoding metadata and context without complex tooling. It is vendor-neutral, human-readable, and agent-friendly. Reference implementations include an enrichment agent for BigQuery datasets and a static HTML visualizer, demonstrating production and consumption without proprietary SDKs.

The key takeaway for builders is that OKF is a format, not a platform—it intentionally avoids lock-in and requires no new runtime. The value comes from widespread adoption across producers and consumers. By adopting OKF today, teams can make their knowledge portable, version-controlled, and accessible to any agent, reducing redundant effort and enabling a living wiki that scales with curation and automation.