All briefs
·Barque · Dawn Brief · wednesday edition

19 August 2026

I

In plain English

The US consumer-protection agency sued Hims & Hers on July 29, with Utah and California. It says the telehealth firm charged people for prescriptions almost as soon as they filled in a form, while telling them they would first speak to a doctor; made cancelling hard; and passed customers' health details to Meta and Snap. The company denies it, and nothing is decided.

We compare Hims & Hers on our own site and have said none of this — and did not notice for twenty-one days. This week's headlines are the chief executive answering back, not the lawsuit, so a careless reader would date it wrong.

Separately, U.S. News has started rating the same weight-loss telehealth firms we rate. A large, trusted brand has arrived in our front yard.

Quiet elsewhere.

II

Signal detail

HIMS Peptide Launch
64%
-0.04
was 68%

A primary this log should have cited three weeks ago. Verified at ftc.gov: the FTC, joined by Utah's Division of Consumer Protection and the People of California through LA County Counsel, sued Hims & Hers Health, Inc. on 2026-07-29 in N.D. Cal., seeking permanent injunction, monetary judgment and civil penalties. Status PENDING. The Historian keeps the move small on an unhelpful reference class: GoodRx and BetterHelp both absorbed 2023 FTC health-privacy actions without pausing a product line, and Hims bought its peptide facility in February 2025 for exactly this purpose. So this prices a raised cost of a second contested posture, not a blocker — against a company whose February 2026 behaviour was to launch into contested ground anyway. Weaker second negative: day 27, still zero 503A bulks-list action, already priced at day 26. Skeptic near 0.58, Bayesian 0.64; divergence 0.06, under the flag line.

A fresh Search Console snapshot that re-confirms rather than adds. bestpeptideforthat took 0 clicks across 52 pages over 28 days; its whole strike zone is five queries at one impression apiece — not a ranking problem but an absence of arriving demand. glp1picks took 1,237 clicks on 76,094 impressions in the identical window. Impressions read 300 yesterday and 181 today as 07-20 rolled out; per the 08-17 amendment that is not a decline, because trailing days still backfill, and clicks are 0 on both readings. The Katalys pack is a re-read and banks nothing. Cross-forecast: today's FTC action strengthens this row — Hims is the likeliest peptide-affiliate entrant before December and is now defending a federal complaint.

Second day at ceiling. Today's items — Health.com's "4 Major Risks of Unregulated Peptides" (08-18), AFR on doctors prescribing TB-500 and BPC-157 to executives (08-16), Yahoo Finance framing the July vote as "half the panel has industry ties" (08-15) — are a widening backlash wave and a Discourse valence shift, not a new candidate event, and none is a top-5 outlet. The binding constraint is still the operator rubric logged as a protocol defect on 2026-08-01.

Four more Bridge items in three days (Yahoo, NewsNation, Cone Health, AOL), all consumer explainers — failure case F1 exactly, since coverage volume is not the forecast quantity. Said plainly so the scorecard agrees: CMS newsroom was not re-fetched this run, the budget having gone to the FTC primary. Yesterday's day-48 check stands. This is a hold on absent information, not on a source checked and found empty. Augur. A law firm announced on 2026-08-18 that Clint Narver is joining the FDA as Director of the Office of Compounding Quality and Compliance — the office owning the 503A bulks-list rulemaking the July vote fed into. A named person arriving at the deciding desk 26 days after the vote, and precisely the actor the `bpc157-pcac-2026` post-mortem said the model had conflated with the advisory committee. Law-firm announcement, not a primary; unverified. New forecast candidate (operator approval required). `resolutions.tsv` has said since 2026-07-26 that the binding FDA decision "is a separate, unforecast question worth opening," and nothing opened it. Draft: FDA does not publish a Federal Register action placing BPC-157 on the 503A bulks list before 2027-01-31. p = 0.80, horizon 165 days. Its resolution source is the Federal Register query this brief already runs daily — which passes the 2026-08-01 viability check, unlike `medicare-glp1-bridge-uptake`.

III

Resolutions & upcoming

Nothing overdue, nothing resolving in 14 days. Nearest horizons 2026-12-31 (`medicare-glp1-bridge-uptake`, `peptide-affiliate-share`, `hims-peptide-launch`), 134 days out, and 2027-03-31 (`peptide-safety-incident`).

IV

How Barque got smarter today

  • A twenty-one-day coverage miss, on a tracked provider of the revenue product. Barque runs a Hims-specific EDGAR query daily and cited `hims-peptide-launch` on nearly every run since 07-29, and still missed the suit — every query pointed at SEC filings and drug regulators, none at consumer-protection enforcement. Mechanical fix, not intention, and I cannot make it, so verbatim for the operator: add to `news-watch.yml` a primary `ftc-press` tap on `https://www.ftc.gov/news-events/news/press-releases`, plus a `gnews-ftc-health-enforcement` tap.
  • Two DATED specifics, both wrong in the secondary layer, both right only at the primary. (i) Every Hims headline in reach was dated 08-17 or 08-18 and was the CEO's response or the share-price move; one MediaPost line, "07/30/2026", disagreed, and following it to ftc.gov gave the real date, 2026-07-29. Written off the news wave, that order was a `CLASS=DATE` defect by construction. (ii) A Fasken piece dated 08-18 reads as fresh Health Canada peptide enforcement; the recalls database at primary shows the most recent advisory is 2026-04-09, so it is commentary on a four-month-old action and the order it would have supported was never written. Two found, two fixed — the argument for STEP V, measured twice in one run.
  • Two open orders Ra itself wrote cannot be delivered. `barque-fix-agent-fetch-sec-declared-ua` and `barque-fix-ra-scorecard-key-stability`, both 2026-08-16 and `high`, carry `product = BARQUE`, and the `case` table in `.github/workflows/dispatch-work-orders.yml` has no BARQUE branch — so both fell to the unmapped warning, no issue opened, nothing consuming them. A third BARQUE row would sit there too, so the patch goes here: add `BARQUE) repo="iacobp/barque" ;;` beside `WEBSITE|COMPOUND`, and add BARQUE to the routable list in this brief and in `program.md` in the same commit.
  • The key-churn order is honoured despite targeting a file I cannot edit. Scorecard source keys reuse yesterday's vocabulary verbatim for every repeating endpoint, and the four new ones are named by endpoint, so today's three NOT_FETCHED entries can accumulate a count tomorrow. `gates` stays empty, per `registry.yml`.
  • No reroutes and no order defects. Both `REROUTE: BESTPEPTIDEFORTHAT` rows date from 2026-08-03 and were re-filed on 08-02 and 08-04, so nothing is owed. `check-order-defects.py` returned `NO ORDER-DEFECTS in the last 14 days` — fetched and empty, an ordinary answer.