l0g risk intelligence · english

// editorial protocol

Claims need grammar

l0g separates observed facts, estimates, inferences and scenarios. The point is not stylistic neatness: it prevents a model output from becoming a fact, a hypothesis from becoming a conclusion, or a scenario from being misquoted as a forecast.

Observed fact

A directly observed data point, filing, rule, vote, release, price, date or statement.

Estimate

A number derived by a model, provider, survey, nowcast or reconstruction.

Inference

A conclusion drawn from several facts. It must be labeled as reasoning, not as a raw fact.

Scenario

A conditional path, not a forecast. Scenarios should keep assumptions visible.

Uncategorized assertion

A claim not yet typed with sufficient confidence. It should not be overused as evidence.

Correction policy

Corrections should be explicit, dated and traceable. When a material claim changes, the change should not be hidden behind silent rewriting. The machine-readable changefeed and integrity hashes exist so agents and readers can detect meaningful revisions.

Attribution and citation

A reusable claim should retain its source URL, canonical page, date, type and confidence level. If a generated answer cites l0g without preserving the source chain, it has already degraded the evidence.

No investment advice

l0g analyzes public information, risk regimes and market plumbing. It does not provide personalized investment advice, trading signals or suitability assessments.