agenticweb.wiki

Open Knowledge Format

Last checked · 17 primary sources

This page is about the Open Knowledge Format (OKF), Google Cloud's Markdown format for knowledge bundles that people and AI agents read. It has nothing to do with the Open Knowledge Foundation (okfn.org), a nonprofit that also uses the initials OKF [17]. For a site-level list of pages for LLMs, see llms.txt. For instructions to coding agents in a repository, see AGENTS.md.

Overview

The Open Knowledge Format (OKF) is an open specification from Google Cloud's Data Cloud team for writing down an organisation's knowledge, such as table schemas, metric definitions and runbooks, as "a directory of markdown files with YAML frontmatter" [1][5]. Each file is one concept and type is the only required frontmatter field. Consumers must not reject unknown fields, concepts link to each other with ordinary Markdown links, and any directory may hold optional index.md and log.md files [1]. Version 0.2 (July 2026) added optional fields for provenance (sources), trust (generated, verified) and lifecycle (status, stale_after), plus a new Attested Computation concept type [1][6]. Google says OKF "formalizes the LLM-wiki pattern", which it says already shows up in Obsidian vaults used by coding agents, AGENTS.md and CLAUDE.md files, and repositories of index.md and log.md files [5].

Who uses it

Each URL below is the root index.md of a published OKF bundle. On 2026-10-28 each one returned HTTP 200 with no redirect, began with YAML frontmatter and declared the okf_version shown. This does not certify that every concept file in the bundle conforms to the spec. Google's own sample bundles and agenticweb.wiki are not listed.

Who reads it

No statement found as of 2026-10-28 from a major AI agent or coding-agent vendor (for example Anthropic, OpenAI, Cursor or GitHub) saying its product reads OKF bundles natively.

Recent coverage

See also

References

  1. Google Cloud, Open Knowledge Format (OKF), Version 0.2 (SPEC.md, GoogleCloudPlatform/open-knowledge-format, main branch), checked 2026-10-28 (text checked at the raw.githubusercontent.com copy): "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." "`type` is the only always-required key; a concept carrying just `type` is fully conformant (§11)." "Consumers SHOULD preserve unknown keys when round-tripping and MUST NOT reject documents with unrecognized fields." "A bundle MAY be distributed as: - A git repository (recommended, since it provides history, attribution, and diffs). - A tarball or zip archive of the directory. - A subdirectory within a larger repository." "OKF v0.2 makes provenance, trust, lifecycle, and attestation first-class while keeping the format minimally opinionated." "Replacing domain-specific schemas (Avro, Protobuf, OpenAPI, and so on)." — https://github.com/GoogleCloudPlatform/open-knowledge-format/blob/main/SPEC.md
  2. Google Cloud, GoogleCloudPlatform/open-knowledge-format (GitHub repository README; GitHub repository metadata shows it was created 2026-08-11 and is licensed Apache-2.0), checked 2026-10-28: "Bundles live in git. Pull requests, line-by-line diffs, blame, and review workflows just work" — https://github.com/GoogleCloudPlatform/open-knowledge-format
  3. Google Cloud, Open Knowledge Format (OKF) (okf/README.md, GoogleCloudPlatform/knowledge-catalog: the old location), checked 2026-10-28: "That repository is the canonical home of the specification, the reference agent, and the sample bundles." "It is a frozen snapshot, no longer maintained, and anything built against it will drift out of date." — https://raw.githubusercontent.com/GoogleCloudPlatform/knowledge-catalog/main/okf/README.md
  4. GitHub, commit history of okf/SPEC.md in GoogleCloudPlatform/knowledge-catalog (GitHub API): v0.2 commit 2026-07-24; timestamp amendment 2026-08-21: "okf: migrate format and tooling to Open Knowledge Format v0.2 (#227)" "okf: make every timestamp an ISO 8601 datetime with an explicit offset (#323)" — https://api.github.com/repos/GoogleCloudPlatform/knowledge-catalog/commits?path=okf/SPEC.md&per_page=100
  5. Sam McVeety and Amir Hormati, Google Cloud, How the Open Knowledge Format can improve data sharing (Google Cloud Blog; on-page headline "Introducing the Open Knowledge Format"), 2026-06-12: "That's why today, we're introducing the Open Knowledge Format (OKF), an open specification that formalizes the LLM-wiki pattern into a portable, interoperable format." "OKF requires exactly one thing of every concept: a type field." "These are proofs of concept, deliberately." "OKF v0.1 is a starting point, not a finished standard." "We have also updated Google Cloud's Knowledge Catalog to be able to ingest Open Knowledge Format and serve it to our agents." "Published by the Google Cloud Data Cloud team." "Similar knowledge-as-Wiki pattern keeps reappearing under different names: Obsidian vaults wired to coding agents, the AGENTS.md / CLAUDE.md family of convention files, repos full of index.md and log.md artifacts that agents consult before doing real work, and "metadata as code" repositories inside data teams." — https://cloud.google.com/blog/products/data-analytics/how-the-open-knowledge-format-can-improve-data-sharing
  6. Sam McVeety and Amir Hormati, Google Cloud, OKF v0.2 adds trust signals (Google Cloud Blog), 2026-07-24 (page metadata; the page as fetched shows 25 July 2026): "In OKF v0.2, it is now possible to answer all five of those questions from frontmatter, while the format remains as minimally opinionated as v0.1." — https://cloud.google.com/blog/products/data-analytics/okf-v0-2-adds-trust-signals
  7. Firat Elbey and Vasiya Krishnan, Google Cloud, Knowledge Catalog: self-learning context enrichment for AI agents (Google Cloud Blog), 2026-10-22 (page metadata; the page as fetched shows 23 October 2026): "because context sets build on the Open Knowledge Format (OKF), you can import an OKF bundle to become a context set and export it back out as OKF so your context stays portable." — https://cloud.google.com/blog/products/data-analytics/knowledge-catalog-self-learning-context-enrichment-for-ai-agents
  8. atomicstrata, Release v0.10.0 · atomicstrata/llm-wiki-compiler (GitHub release notes), 2026-06-14: "Adds a review-policy gate for generated knowledge, a full Open Knowledge Format (OKF) export/import round-trip, and a Mintlify documentation site." — https://github.com/atomicstrata/llm-wiki-compiler/releases/tag/v0.10.0
  9. Rodrigo Serradura, Your bundles, served to any agent: okf-mcp 1.0.0 (okf-gem blog), 2026-08-11: "okf-mcp 1.0.0 adds the surface that needed none of that: any host speaking the Model Context Protocol, Claude Desktop, Claude Code, or whatever your team runs next, reads the bundles on your machine directly." — https://okfgem.com/blog/okf-mcp-1-0-0/
  10. Memanto, Open Knowledge Format (OKF) - MEMANTO Docs, page metadata last modified 2026-10-19: "Memanto exports OKF v0.2 bundles; the bundle's root index.md declares okf_version: "0.2"." "Load any OKF bundle into an agent with memanto migrate okf" — https://docs.memanto.ai/integrations/okf
  11. Danielmelody (GitHub user), Community implementation: xMatrix bidirectional OKF Pages (issue #37, GoogleCloudPlatform/open-knowledge-format), 2026-10-12 (text checked through the GitHub API): "xMatrix now supports bidirectional OKF v0.2 bundles for its collaborative Pages." "This is a community implementation, not a request to imply Google endorsement." — https://github.com/GoogleCloudPlatform/open-knowledge-format/issues/37
  12. Aembit, AI Assistant Integration Resources (Aembit docs), no date shown; checked 2026-10-28: "Aembit publishes its documentation in two machine-readable formats for AI coding assistants and large language models (LLMs)." "Use the OKF bundle when your agent has a persistent workspace and file tools (Claude Code, Cursor, Copilot agents)." — https://docs.aembit.io/llm-resources/
  13. The Website Specification, Open Knowledge Format (OKF) bundle, no date shown; checked 2026-10-28: "This site ships it: the bundle targets OKF 0.2, is generated from the same content collection as every other surface, and is served browsable at /okf/, with the whole tree packaged as /okf.tar.gz." — https://specification.website/spec/agent-readiness/okf-bundle/
  14. Matthias Bastian, The Decoder, Google Cloud's Open Knowledge Format turns scattered docs into Markdown files for AI agents, 2026-06-14: "Google Cloud is introducing the Open Knowledge Format (OKF), a new spec that standardizes knowledge as Markdown files and makes it portable across systems." — https://the-decoder.com/google-clouds-open-knowledge-format-turns-scattered-docs-into-markdown-files-for-ai-agents/
  15. Asif Razzaq, MarkTechPost, Google Cloud Introduces Open Knowledge Format (OKF): A Vendor-Neutral Markdown Spec for Giving AI Agents Curated Context, 2026-06-16: "OKF is a format, not a service or a platform." — https://www.marktechpost.com/2026/06/16/google-cloud-introduces-open-knowledge-format-okf-a-vendor-neutral-markdown-spec-for-giving-ai-agents-curated-context/
  16. Janakiram MSV, The New Stack, Andrej Karpathy, Google and Garry Tan agree Markdown is the answer, but they're not solving the same problem, 2026-07-06: "If no one develops consumers for it, it remains just a good idea that Google released on a slow Friday." — https://thenewstack.io/markdown-agent-memory-moat/
  17. Open Knowledge Foundation, Open Knowledge Foundation – For a fair, sustainable and open future (okfn.org home page), checked 2026-10-28: "Open Knowledge Foundation – For a fair, sustainable and open future" — https://okfn.org/en/

JSON · Markdown · OKF v0.2