Run every canary job weekly on Mondays

.github/workflows/canary.yml

@@ -8,16 +8,15 @@
 # datacentre address, listings and product pages answered HTTP 200 to a plain
 # `curl`, to `python-requests` and to a request with no User-Agent header at
 # all, with no challenge of any kind. A GitHub runner is a datacentre address
-# too, so this workflow re-measures that claim every morning.
+# too, so this workflow re-measures that claim every Monday.
 #
 # **A canary that CAN pass without a credential must never be gated on one.**
 # If Kosé ever puts its catalogue behind a key or a bot manager, this badge
 # goes red the next morning and the README's claim is retested without anyone
 # remembering to.
 #
 # The last job exercises the PAID path (the Scraping Browser API over CDP),
-# which cannot run without a secret, and bills a session, so it runs on the
-# weekly (Monday) schedule and on a manual dispatch only. It does not pass by skipping: with no
+# which cannot run without a secret. It does not pass by skipping: with no
 # secret it FAILS with an ::error:: naming the secret, because a green check
 # that tested nothing reads exactly like one that tested something.
 #
@@ -32,9 +31,6 @@ on:
   schedule:
     # 05:40 UTC. Not on the hour: every scheduled workflow on GitHub fires at
     # :00 and the queue there is long.
-    - cron: "40 5 * * *"
-    # Mondays only: the paid job (scraping_browser) runs on this
-    # schedule (each run bills); the free jobs keep the daily one above.
     - cron: "40 5 * * 1"
   workflow_dispatch:
 
@@ -381,8 +377,6 @@ jobs:
 
   scraping_browser:
     name: Scraping Browser API (needs the KOSE_CDP_ENDPOINT secret)
-    # Paid path: weekly schedule or a manual dispatch only.
-    if: github.event_name != 'schedule' || github.event.schedule == '40 5 * * 1'
     runs-on: ubuntu-latest
     timeout-minutes: 20
     env:

CHANGELOG.md

@@ -12,9 +12,7 @@ existing user would notice, the release notes lead with it.
 
 ### Changed
 
-- CI only: the canary's paid `scraping_browser` job now runs weekly (Mondays
-  05:40 UTC) and on a manual dispatch; the free `listing`, `http_vs_browser` and
-  `product` jobs still run daily.
+- CI only: the whole canary, free and paid jobs alike, now runs weekly (Mondays 05:40 UTC) and on a manual dispatch; the daily schedule is removed.
 
 ## [0.3.0] — 2026-10-15
 

README.md

@@ -249,7 +249,7 @@ So until this repo had a browserless engine, every long title came back
 short — **10 of 72 rows** on one live 3-page run — and nothing could see it,
 because there was nothing to disagree with. The parser now recovers the full
 name from the thumbnail's `alt`, which the rewrite does not touch and which
-is present on 171 of 171 measured tiles. The canary runs both engines daily
+is present on 171 of 171 measured tiles. The canary runs both engines weekly (Mondays)
 and fails if any field differs.
 
 **The brand in a URL selects nothing.** `/site/xyz/c/c15/` returns

smoke_test.py

@@ -2156,14 +2156,14 @@ def check_captcha_capability_claims_match_the_code():
     # rots the moment the site switches a bot manager on.
     #
     # So the claim has to be PAIRED with the thing that would retest it — the
-    # daily canary — rather than left as a sentence nobody re-measures.
+    # scheduled canary — rather than left as a sentence nobody re-measures.
     claims_no_key = ("no, and it would be dishonest" in low
                      or "do i need a 2captcha account" in low)
     if claims_no_key:
         check("README pairs 'no key needed' with the canary that retests it",
               "canary" in low,
               "a claim that the site is ungated goes stale silently; the "
-              "daily canary is what turns it back into a measurement")
+              "scheduled canary is what turns it back into a measurement")
         check("...and dates the measurement it rests on",
               re.search(r"20\d\d-\d\d-\d\d", readme) is not None,
               "a number describing a living thing needs its moment")