The journey passed whenever the LA map was empty or not yet rendered. It now
waits for a state school in the same LA to show its gap, and both LA journeys
fail rather than skip on an empty map.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A '2025-26-revised' release read as an unknown year went last, behind the
'2025-26' release, which then owned the year and dropped every revised row.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The plan hard-coded one model's Co-Authored-By trailer; an executor should use the one its own session specifies. The spec's report-card rule now matches int_ofsted_latest. Review on #183.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
int_ofsted_latest treated any report card as the latest visit, relying on report cards postdating legacy inspections. That holds today (0 of 2,451 report-card schools in the 31 Aug 2026 MI) but is now not assumed: the latest visit is the newest of the three dates. A report card still leaves no legacy grade in force, which is the spec's rule rather than an inference. New unit test newer_legacy_visit_after_a_report_card failed RED on the old logic. Review on #183.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Rabbsfarm (102408) must read 'Inspected · 2025' with no overall grade, and Robins Lane (104762) must say its Good was confirmed at an ungraded inspection on 18 July 2024. Both fail against staging today (no current_grade field, no confirmation line) and pass once PR #183's pipeline run and this branch are deployed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
buildOfstedHeroChip and buildSchoolSummary were rendered nowhere and encoded the old carried-forward rule (and a framework value, 'ReportCard', the API never sends).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
'Inspected' is the school's latest visit, so Washwood Heath reads 21 May 2025 rather than '3 Mar 2020 · 4+ years ago' (audit M1). The Result row names the inspection a grade came from, or says 'No overall grade'; 'Grade carried forward' and 'transitional framework' go.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The section is dated by the latest visit and headlines the grade still in force with the inspection that awarded or confirmed it, or 'No overall grade'. A later visit that isn't the grade's source gets its own line, with an ungraded outcome such as 'Standards maintained'. The primary and secondary no-grade branches merge, so the secondary page no longer drops the sixth-form and early-years judgements (audit M2), and 'Not rated' goes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Search badges date a grade by the inspection that awarded or confirmed it (ofsted_grade_date) and say 'Inspected · year' when the latest inspection gave no grade. ofstedDisplay's kinds become graded / confirmed / no_overall_grade, read from current_grade rather than a carried-forward overall_effectiveness. lib/ofstedStatus.ts holds the two sentences the school and compare pages share.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The list query and the batch Ofsted fetch each picked 'the latest row' of fact_ofsted_inspection themselves, and _ofsted_block carried an older ungraded grade forward when the latest graded inspection gave none. All three now read marts.fact_ofsted_latest. List rows: ofsted_grade is the grade still in force, ofsted_grade_date when it was awarded or confirmed, ofsted_date the latest visit. The ofsted block gains current_grade and latest_visit and loses grade_source; overall_effectiveness is the graded inspection's own result. A school with only an inspection stays publishable in the sitemap.
Requires fact_ofsted_latest (PR #183's pipeline run) on the database.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
dim_school.ofsted_grade (Typesense's rating) coalesced the graded grade with an older ungraded visit's, so Rabbsfarm's 2020 'remains Good' survived its 2025 no-grade inspection. It now reads int_ofsted_latest's current_grade, and ofsted_date is the latest visit. Unit tests pin both (audit C1, M1).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The backend will read this instead of picking the latest row of fact_ofsted_inspection itself (twice, with arbitrary ties). Schema tests and assert_ofsted_current_grade_consistent pin its invariants. Built by the monthly Ofsted DAG (int_ofsted_latest+), not the daily one.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
int_ofsted_latest now picks each school's latest visit (report card, graded or ungraded inspection) and the overall grade still in force, dated by the inspection that awarded or confirmed it. A 'School remains Good' from an older ungraded visit is no longer carried past a newer inspection that gave no grade (audit C1). Duplicate monthly rows resolve to the newest visit, so a newer report card always wins. dim_school's columns are unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
int_ofsted_latest needs to know which inspection came last. Report-card-only schools are no longer dropped (123 schools, audit H3); inspection_date stays for current readers and now falls back to the report-card date, so fact_ofsted_inspection's not_null test still holds.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Design for audit findings C1 and M1. A grade carried forward from an older
ungraded visit is shown under a newer inspection's date ("Good · 2025" for
Rabbsfarm, whose 2025 inspection gave no grade), and "Inspected" dates show the
last graded inspection rather than the latest visit. The rule moves into
int_ofsted_latest and a new fact_ofsted_latest mart, shipped as a pipeline PR
and then a backend/UI PR.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
GIAS publishes its CSVs in Windows-1252 and sends no charset. The tap read
resp.text, so requests guessed the codec, and the encoding="latin-1" passed
to read_csv did nothing on already-decoded text. On 3 Oct 2026 the guess was
windows-1250, and "St Thomas à Becket" (138950, 149557) was stored as
"St Thomas ŕ Becket". A different guess on another day would garble other
accented names.
Both streams now decode the downloaded bytes themselves (gias_csv.py). A byte
Windows-1252 leaves undefined becomes U+FFFD with a logged warning instead of
failing the load, so one odd name cannot stop the daily refresh. None of the
nine extracts checked (1 Jul to 3 Oct 2026) contains such a byte.
Checked by running the tap on the real 3 Oct extract with .text forced to
windows-1250: all 52,586 rows decode, with no "ŕ" and no replacement
characters.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
fact_ks4_destinations and fact_ks5_destinations join dim_school, so the
daily build's stg_gias_establishments+ and the monthly Ofsted build's
dim_school+ both selected them. They also read stg_ees_ks4/ks5_destinations,
which only the manually triggered EES DAG builds. Where that DAG hasn't run
since the destinations models landed, dbt_build fails with "relation
staging.stg_ees_ks4_destinations does not exist", and sync_typesense and
invalidate_cache never run. Production's register data has been stuck at
about 25 Aug 2026.
Both builds now exclude the descendants of the two EES staging models, as
the daily build already does for the KS2/KS4 lineage models. The EES DAG
still rebuilds the marts when their data changes.
test_dag_selectors reads the model graph from the SQL (CI has no dbt) and
checks that every scheduled build only reads models it or the daily build
builds. It failed for the daily and monthly Ofsted builds before this change.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
"Other schools nearby" was followed by "Other primary schools near
<school>.", which says the same thing again. The heading now stands alone.
nearbyNoun() and the phase and schoolName props existed only to build that
line, so they go with it. Its bottom margin was the only gap between the
heading and the cards, so the header row carries that gap now, and centres
the heading against the carousel arrows now that it is a single line.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The Admissions section of a non-selective secondary with no religious
character must say neither "Selective:" nor "Faith priority:". Run against
staging before the fix, it fails on "Faith priority" (Burntwood's "(None)").
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The header's Selective flag read school_info.admissions_policy, which the
detail endpoint never sent, so no school page could flag Selective while
its search row did (staging E2E: The Grammar School at Leeds). The detail
payload now carries it, and a contract test checks it carries every field
the header's flags read.
Sending it would have switched on two older copies of the tag logic #176
fixed in the rows. The Admissions section and the cut-off note both tested
includes('selective'), so every non-selective secondary would have read
"entry is by selective examination". The section also counted "None" as a
faith: Burntwood reads "a faith-based admissions priority (None)" today.
All of them now share isSelective() and hasReligiousCharacter(), which also
treats "Not applicable" as no faith, as the place table already does.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The search page fetched LA averages with cache: 'force-cache', which serves
any stored response, however old, without asking the server. One failed
request (a staging deploy restart; the July proxy outage) was stored and
replayed on every later visit, and the error was swallowed, so the
"vs LA avg" delta silently vanished from every secondary row in that
browser. A Playwright profile still held a 500 dated 5 July.
The default cache mode honours the API's Cache-Control (five minutes), so
a good answer is still reused and an error never is. Browsers holding a
stored failure recover on their next visit.
A journey now checks that a mainstream secondary's row shows the
comparison: nothing did, which is how it could go missing unnoticed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
tudor
