• Probability, not timing. Strategic warning is a judgment that significant action is probable — not a forecast of when. Indications often thin out just before action.
• Totality of evidence. No single indicator is warning. Each observation only shifts the probability.
• Facts vs. indications. A fact becomes an indication only in context.
• Capability · Intent · Preparations. The three questions; sub-indicators carry the lens they inform.
• Conveyance. Warning is not complete until the decision-maker knows they have been warned.

Grounded in Cynthia M. Grabo, Anticipating Surprise: Analysis for Strategic Warning, and “Strategic Warning: The Problem of Timing” — CIA Center for the Study of Intelligence & DTIC.

• Indicator (I-n). A named category of activity we would expect to see if a warning problem were developing — e.g., I-4 Conflict & Stability. Its status is the worst of its sub-indicators.
• Sub-indicator (S-n.m). One specific, collectable slice of an indicator, bound to exactly one observable and one open source — e.g., S-4.1 Conflict media velocity ← GDELT.
• Observable. The phenomenon a source could measure, defined in advance together with its condition — e.g., "protest/conflict-coded article volume in the AOI over 24h, threshold ≥20."
• Observation. What the source actually returned at assessment time — a measured value with a timestamp. In Grabo's terms, a fact.
• Trigger. The event of an observation crossing its stated condition. The sub-indicator "fires," and the fact becomes an indication — meaningful only in context, weighed against the totality.
• Judgment. The probability posture rolled up from all indications — domain-weighted: I&W indicators (Conflict, Cyber, Air, Maritime) drive the posture; situational hazards and context inform it at reduced weight, and the phrase states when hazards rather than indications are the driver. Confidence scales with sources reporting, and low confidence is flagged as provisional. Never a forecast of timing.

The console's registry — every indicator and sub-indicator, numbered, with its observable, source, and wiring status. Unwired rows are honest gaps: candidates awaiting a keyed or future source.