Knowledge inventory & concept mapping
A structured audit of everything your organization can claim, teach, or prove, broken into the discrete concepts a bundle is built from.
OKF Creation
Publish your knowledge in the format AI agents read natively.
The Open Knowledge Format, published into the Google Cloud ecosystem in June 2026, packages what your organization knows as structured markdown bundles: one concept per file, YAML facts on top, human explanation below. Agents parse it directly instead of scraping your site and guessing. We inventory your knowledge, author the bundle, and keep it in sync. Almost nobody offers this yet. That's precisely the point.
Jun 2026 the OKF spec went live. The format is weeks old, and the early-mover window is open now
Zero ranking benefit promised today. OKF is an agent-readability bet, and we say so plainly
1 sprint to inventory, author, and publish your first bundle. It compounds with every content update after
Why it matters now
Search asked for pages.
Agents ask for knowledge.
The next buyer research loop isn't a person typing queries; it's their AI agent gathering facts to compare options. Agents work best with structured, declared knowledge, and OKF is the first open format built for exactly that handoff. Nothing crawls the open web for bundles yet, and honest practitioners admit it. But formats reward the early: the brands whose knowledge is already structured when agent traffic arrives get read first, and structuring your knowledge now costs a sprint. Retrofitting it under pressure later costs a quarter.
Map what you actually know
Services, products, methods, pricing logic, FAQs, and proof points inventoried as discrete concepts. Most organizations have never listed what they know; the inventory alone sharpens every other channel.
Structure it to spec
One concept per file: YAML frontmatter carrying the machine-checkable facts, markdown body carrying the explanation. Written to the current spec, validated, and versioned.
Publish and keep it true
The bundle deployed on your domain, tied into llms.txt and your schema layer, and updated when your facts change. A stale knowledge file is worse than none.
What's included
From scattered expertise to a versioned bundle.
Bundle authoring
Every concept file written to spec: clean frontmatter, accurate claims, readable explanations. Authored by people who write for machines daily, not exported by a script.
Entity alignment
Bundle claims kept consistent with your schema markup, Wikidata item, and site copy. Machines cross-check; contradictions cost trust everywhere at once.
Publishing & hosting setup
The bundle deployed on your domain with discovery signals in place: llms.txt reference, sitemap inclusion, and clean URLs agents can fetch.
Sync workflow
A lightweight process tying bundle updates to your content calendar, so new services, prices, and proof points land in the bundle the week they change.
Spec-watch & migration
The format is weeks old and will evolve. We track the spec and migrate your bundle as it matures, so the early bet stays current.
How the work runs
One sprint to publish. Then it lives with your content.
First
Inventory the knowledge
Workshops and content mining produce the concept map: what you know, what you can prove, and what an agent would need to recommend you.
Author
Write the bundle
Concept files drafted, fact-checked against your actual offering, and validated against the spec. You review every claim before it ships.
Publish
Deploy with discovery
The bundle goes live on your domain, referenced from llms.txt and aligned with your schema, ready for whatever starts crawling first.
Maintain
Sync and migrate
Facts stay current through the sync workflow; the format stays current through spec-watch. The bundle ages like infrastructure, not like a campaign.
Client Outcome
What early structural bets look like when the traffic arrives.
From the results library: measured performance across a full engagement window.
25+ local-intent searches held in the top 5, against the maps pack, aggregators, and every academy in the city.
4,890 organic clicks in 90 days
341K search impressions
Top 5 for 25+ local intent queries
Read the full story“We used to be invisible unless someone typed our name. Now we get found by people who didn't know we existed. The enquiries come in already knowing what class they want.”
Straight answers
Questions we
get asked first.
No sales spin. If the honest answer is "it depends," we tell you what it depends on.
- What is OKF?
- The Open Knowledge Format is an open specification, published in the Google Cloud ecosystem in June 2026, for packaging organizational knowledge as a bundle of markdown files. Each file describes one concept, a service, a method, a dataset, an FAQ, with structured YAML facts on top and a human-readable explanation below, so AI agents can parse your knowledge directly instead of scraping and guessing.
- Will an OKF bundle improve my rankings?
- Not today, and anyone claiming otherwise is selling hype. OKF is not a known ranking signal and nothing crawls the open web for bundles yet. The case is different: it's a low-cost early position for agent-driven discovery, and the inventory work sharpens your schema, your llms.txt, and your site copy immediately.
- Who should create an OKF bundle now?
- Brands already investing in entity work and GEO, where the marginal cost is one sprint and the knowledge inventory pays for itself in content clarity. If you haven't fixed rendering, schema, or your entity graph yet, do those first; we'll tell you so in the scoping call.
- How is OKF different from llms.txt?
- Different jobs. llms.txt is a signpost: a short file telling AI systems what your site covers and where to look. An OKF bundle is the payload: the actual structured knowledge, concept by concept, in a format built for machine consumption. They work best deployed together, and we ship them aligned.
- What happens if the OKF spec dies?
- You keep the assets. The knowledge inventory, the structured claims, and the entity alignment all feed your schema, your content, and any future format directly. The bundle is one export of work that pays regardless, which is what makes this an asymmetric bet rather than a gamble.
Ready when you are
Be the brand whose knowledge is already structured.
Free scoping call: what your bundle would contain, what it costs to build, and an honest read on whether you should do it now or fix deeper layers first.
Scope your OKF bundle