An organizational world model, built deliberately.

Preloaded knowledge

The World Intelligence Core contains ten domains, twenty capabilities and forty Work Types. Selected educational pages expose their proposed criteria, evidence expectations and source-supported statements. Operating defaults remain L0 hypotheses; independent domain review and empirical validation are pending. The Role Library adds six Manager and twenty Worker specifications. Selected educational Role pages remain unvalidated hypotheses; the original ten Role identities and prior release remain preserved.

Private operational knowledge

Customer objectives, attempts, decisions and evidence remain organization-scoped. Public canonical records are loaded from a separate release artifact. No customer data is pooled or published.

Evidence before conclusions

A useful intelligence system must expose source, transformation and uncertainty. The evaluation harness tests synthetic scenarios with deterministic probes; it does not establish live AI capability. No customer-learning pipeline or economic benchmark is active.

How well do we know?

Hypothesis means a proposed default. Researched means a scoped assertion has supporting sources. Evaluated means a claim has controlled-test evidence. Observed requires permitted real organizational execution. Outcome-validated requires independently verified resulting state. These levels are distinct from editorial publication.

Current Role definitions are hypotheses. Source-supported Work statements do not validate an entire workflow. The evaluation foundation measures test-harness behavior; no live-model benchmark corpus or public empirical AI-performance finding exists yet.