Alerts

Wide diagram — the chains above are its mobile view.

What this console is: the correlation layer — the R in XDR. Seven of the fourteen live trackers map to a defensible STRONGMAN technique; those readings, the narrative campaigns, and the automated detections are placed on the objective they indicate and joined into attack-pattern chains with a stated end goal. The other seven are disclosed, not forced. Read the console top-down: what is above baseline now, then the chains, then the framework map. Everything listed is at alert or elevated severity; open a card to see what tripped it — the reading, the other readings it is chained to, and, for a narrative campaign, its full lifecycle.

Chains: curated pattern templates that activate only when every objective on the path shows a present reading, following the framework’s documented phase progression. Each row lists the alert cards it joins. The framework map places each reading on the twelve STRONGMAN objectives. It is supporting context for the alert list above, not a separate finding — every node links to its alert card or its tracker. On a narrow screen the map is hidden and the chains above are its mobile view.

Detection (Narrative Watch): Public US news headlines (GDELT DOC 2.0, 48-hour window) are grouped by shared phrasing across independent outlets. A phrase is flagged when its cross-outlet breadth jumps sharply above its own rolling baseline. Until enough days of history exist the baseline is labeled immature — a surge only means something measured against normal.

Who said it: The phrase is searched against the govinfo Compilation of Presidential Documents (White House statements, remarks, and orders). A phrase with no matching document is reported honestly as media-first. That archive lags publication by roughly six weeks, so a very recent statement may not be indexed yet — the document date is always shown so recency stays visible. We never invent an origin.

Who amplified it: The outlets GDELT reports carrying the phrase are classified by political lean using this site’s own media-landscape dataset (compiled from AllSides, Media Bias/Fact Check, and Ad Fontes). “Right-lean share” is the share of rated outlets with a bias above +1; outlets we do not rate are counted separately, never guessed. Reach is outlet breadth — a proxy for amplification, not a measure of belief. Polling integration is planned.

Where it landed: The phrase is matched (case-insensitive, light stemming) against the titles and abstracts of the executive-orders, abortion-laws, and voting-laws trackers. A match links the record; no match means the narrative has not become policy yet.

Campaign score = outlet spread × surge velocity — how many outlets carried a phrase, times how fast that broke from its baseline. It ranks intensity only; a phrase reaches the alert list when its origin is confirmed to the White House, when it lands in policy, or when it concentrates in the right-wing ecosystem (≥60% right-lean carriers) — not on its score.

The phrase scores above are raw velocity readings. Phrases below the campaign gate are tracked on the watch floor, not the alert list.