We monitored our 1,034-account roster, the anonymous boards we track, and our off-roster finds for the week of August 24 to 30, 2026, reaching 97 of the 152 X handles on the roster. 291 items crossed our line, double last week’s count. 255 of them run 39 distinct scripts and the rest stand alone. For the first time in this record a script reached the top tier: the replacement premise and its remigration remedy, fifteen items scoring 4.03, one claim-script carried by two separate sets with no evidenced operational link between them, anonymous board threads and named accounts around the Save Europe Act European Citizens’ Initiative, in the week the European Commission refused to register that initiative and its organizers pivoted to court threats and a million-signature push. That script continues a chain we have carded since the Channel-invasion ruling three weeks back, and it is the first chain in the record to pick up a documented coordination verdict mid-life. The week’s center of gravity is Ceuta, where, per the Financial Times reporting our trigger scan records, mobs attacked migrant encampments on successive nights: three separate scripts work the aftermath, one celebrating the violence as homeland defense, one casting the arrivals as a rape and crime menace, one recasting the landings as conquest. The same reporting describes attacks organized over WhatsApp; the accounts applauding online show no documented link to the mob or to each other. Reform UK’s leader posted that young males without identity documents who come from “dangerous cultures” should not be free to walk the streets of Essex towns, and a sitting MP of his party echoed the line days later against undocumented young men from “backward cultures” roaming streets and parks, the week’s second documented coordination signal. And on Bluesky, an account called for lethal military force against AfD supporters; we coded it under the same standard we apply in the other direction.
The record doubles: 291 items against last week’s 143, and 14,982,782 platform-reported impressions against last week’s 7.7 million. The comparison is footprint-limited in both directions. The corpus ran wider than any prior window, 10,830 in-window posts against last week’s 9,663, with the Bluesky collection at 4,807 posts and the board collection at 1,589 both larger than before and the research slices read against the full database, but the handle walk reached 97 of 152 roster X handles against 137 last week. Part of the increase is what a bigger corpus surfaces, part of the week’s volume sits unread on the handles we did not reach, and the doubled ledger cannot be fully separated from these collection effects. The severity peak is the record’s first top-tier script: the replacement-and-remigration script at 4.03, fifteen items carried by two separate sets with no evidenced operational link between them: six anonymous board threads, one of which closes on an explicit kill demand, and named carriers around the Save Europe Act European Citizens’ Initiative. The widest reach belongs to the Ceuta celebration script, 3.25 million impressions applauding mob attacks on migrant encampments while the attacks continued. The largest by volume is the Jewish-collective control script at thirty-two items. The anti-Indian script at twenty-nine is the largest of the waves that opened this window, up from three items last week, most of them anonymous board threads that report no view counts, so the script we can measure least is again among the biggest we carded.
Scripts by severity
Click a bar for the full scorecards.
The week's weather, day by day
One dot per real event. Filled dots drew a reaction we confirmed; hollow ones drew none we recorded. Click for the full board.
The 291 items, by harm level
Click for the full ledger.
On the radar
Click a date for the full calendar and the patterns building toward it.
01 · The scripts
Five factors are scored from 1 to 5: reach, toxicity, coordination, mainstreaming risk, and harm potential. Each item is assigned to exactly one script, with overlaps adjudicated on the dominant predicate. Their composite Ember Priority Score is shown on a chart running from 0 to 5; the axis starts at 0 for display, and the score itself cannot fall below 1. Click one for its scorecard. Peak severity this week is 4.03, the first top-tier score in this record, and the median script sits at 2.24. One script stands in the top tier, five in the second, twenty in the third, and thirteen at the bottom. Three scripts carry the severity floor an explicit call for violence sets: the top-tier script exceeds its floor on its own computed score, the anti-Indian script computes into the second tier on its own, and one small board script carrying direct sexual-violence calls sits in the second tier only because of the floor.
02 · The ledger
Click a row for the quote, the coding, and the source.
03 · The week's weather
The events we recorded no reaction to matter as much as the ones we did. Together they show which stories these narratives picked up this week and which drew no reaction our footprint recorded.
04 · On the radar
Nine scripts continue a ruled lineage. The Jewish-collective script is carded for a fifth week, the longest chain in the record, thirty-two items this week, and it moved down a tier, from second to third. The replacement-and-remigration script continues the Channel-invasion chain ruled three weeks back and moved from the third tier to the first; the same chain picked up the record’s first mid-life coordination verdict, overt, on the Save Europe Act campaign. The anti-Indian script continues last week’s carding at twenty-nine items. The Clancy collective-guilt script runs a second week at eighteen items, with the jury due back the Monday after the window closes. The Islam-eliminationist, Muslim-officials, Global-South essentialism, grooming-gang, and Muslim-takeover chains also persist under ruled joins. Thirty scripts are new this week, a count that tracks the doubled corpus more than any change in the weather. Eight of last week’s scripts are not carded this week. We do not read that as decay: a narrative can be absent because it lost its carriers, because it did not clear the bar this week, or because this week’s collection did not surface it, and these records cannot separate those. Each line compares one narrative with itself a week earlier, joined on a key ruled when the narrative was carded rather than on any similarity between the two weeks’ text.
05 · Method
This is a purposive, outcome-selected monitoring sample drawn from a 1,034-account roster spanning 23 ideological categories; this week’s coverage limits are stated below. It is not a random slice of the internet. Roster and board collection is systematic within the coverage limits stated below; off-roster discovery is opportunistic and incomplete, so our off-roster finds show that off-roster carriage exists, not how much of it there is. A hit is language that collectivizes, dehumanizes, or calls for exclusion of a protected group, or that runs dehumanizing or eliminationist framing against a political camp. Criticism of governments, institutions, and named individuals over their own conduct stays protected, however sharp. Every hit rests on a dated source read in full at collection. Where the source was still live, a second, adversarial pass re-verified the quote against it; where it had already gone offline, which covers the 4chan threads in this week’s set, the second pass re-read our collection record and the coding rather than the live page, and the item says so. Items that failed re-verification are not in this report. A separate 15-item depth check, run independently of the pipeline on the largest named-carrier slice, agreed with all fifteen codings. Research and drafting run through Claude under this methodology; every published classification passes the documented verification chain, and Jerm holds final editorial responsibility. One item this week cleared the extremism branch of our gate against a political camp rather than a protected class: a Bluesky call for lethal military force against AfD supporters, coded under the same standard we apply in the other direction. Every other confirmed item names a protected class, or a group our methodology treats as one, including twenty-seven of the twenty-eight violence-tier items, the items coded on the two top harm rungs, glorifying violence and calls for violence; the twenty-eighth is the Bluesky item.
Platform-reported impression counts are capture-time platform counters, not audience counts, and they can rise after we read them. Read each figure as the capture-time reading rather than an audience count.
Reach is not measurable everywhere. The 4chan corpus has no view counts and no persistent handles, so reply and image counts stand in for reach as engagement proxies, and how those proxies relate to actual audience exposure is unknown. Every 4chan thread in this week’s set had already cycled off the live board by the time our verification pass ran, so confirmed for those items means verified against our collection records at capture, not against a second live read.
The X footprint needs stating plainly: the weekly credit budget bound the Saturday catch-up walk, and in-window collection reached 97 of the 152 X handles we track, below last week’s 137. This wider corpus therefore rides on narrower handle coverage, and a narrative living on the handles we did not reach would read here as smaller than it is. Four high-volume accounts truncated at the 200-post cap with their tails not recoverable: JackPosobiec, RadioGenoa, ChinaDaily, and globaltimesnews. The first two are active carriers in this week’s record, so their in-window volume reads low. The daily prefiltered shortlist for August 26 was not written; the database itself holds full coverage of that day and the slice files built from it carry the day, so the gap cost convenience rather than coverage. Ten duplicate captures of the same underlying posts were collapsed at the rulings step, and the ledger counts unique posts. The research cut ran Saturday evening. A reserved Sunday collection run failed partway when the prepaid X credit ran out, reaching 53 posts before the provider began refusing requests; the board and Bluesky tail collected normally, a sweep of everything that landed folds in before publication, and the final hours of Sunday remain uncovered, the standing cost of a Sunday-published weekly. The tail changed little: the anonymous-board wings of the anti-Indian, Jewish-collective, and anti-Black scripts ran at Saturday volume into Sunday morning and are recorded as persistence, and one named Canadian outlet account joined the replacement script by posting its removal slogan verbatim, folded in as the final item.
Coordination scoring leans on shared semantic payload, a shared novel claim, coined term or fabricated detail, except where we can identify an organizing hub. Two scripts carry overt verdicts this week. The replacement-and-remigration script rides a named organizing campaign, the Save Europe Act European Citizens’ Initiative, whose demands are themselves an item in the family and whose two-point program one carrier mirrored inside the refusal week; the verdict rests on that campaign subset alone, no connection between the anonymous-board items and the campaign is documented, and the rulings record says so. The dangerous-cultures script pairs Reform UK’s leader with one of the party’s sitting MPs, two sitting politicians of the same party in near-duplicate phrasing inside one window, “dangerous cultures” against “backward cultures” on the same streets image; we coded it overt on the same organizing-hub logic this record applied to a party’s announced policy in a prior edition. Two more scripts carry covert verdicts, both anonymous-board copypasta campaigns: an identical opening post launched four threads celebrating Palestinian deaths, and near-duplicate copypasta opened repeated threads attributing a surveillance vendor to Jewish financiers. On an anonymous board, one actor and an organized handful cannot be told apart, which is what covert means here. The other thirty-five scripts stay organic parallel. One rule matters most this week: the Ceuta mob organized its attacks over WhatsApp, per the Financial Times reporting our trigger scan records, and that is a fact about the violence, not about the posts. The accounts celebrating the attacks show no documented link to the mob or to each other, so the Ceuta scripts score their coordination on the carrier side alone and stay organic parallel.
The Bluesky lane produced two coded items from 4,807 screened posts. One is the lethal-force call described under the sample and the standard, aimed at AfD supporters from the anti-fascist direction and coded under the same bar as everything else here; the direction of an item changes nothing about the standard it must clear. Two from 4,807 says the lexicon layer matches weakly on that platform rather than that the platform carries nothing; read the low count as a collection limit.
No hit-rate or prevalence claim here may be read as a comparison across ideological lanes. That comparison requires a lane-tier sample for every lane compared, over a common platform footprint. We did not draw that sample this pass.
Five factors, each scored 1 to 5: reach (R), toxicity (T), coordination (C), mainstreaming risk (M), and harm potential (H). They combine as (R·T·C·M1.3·H1.5)1/5.8: multiply all five, with mainstreaming risk raised to the 1.3 power and harm potential to the 1.5 power because those two carry the most real-world weight, then take the 5.8th root. The 5.8 is the sum of the weights (1 + 1 + 1 + 1.3 + 1.5), which brings the result back onto the 1-to-5 scale. Multiplication limits how far any one factor can lift a score, without eliminating its influence. One override runs the other way: an explicit call for violence against the target class sets harm potential to 5, and any script at H=5 takes a P2 floor regardless of its computed score, because an explicit call for violence with no measured reach is still an explicit call for violence. Three scripts carry that floor this week, and the chart marks them; two compute to the second tier or above on their own scores, and one is lifted by the floor alone. P1 starts at 4.0, P2 at 3.0, P3 at 2.0. Coordination scores above the floor only on real signals: a named hub, near-duplicate phrasing, synchronized timing, or a confirmed bridge between actors.