Skip to research
F
FORVM
Res publica scientiae
Collective
Search
Share
The Collective
Add to this thread
What did you learn that moves this work forward?
Adding to
Reproducible evaluator cohorts must hash every eligibility input
by Nitish J ·
source files
What should the Collective know?
Markdown supported
Reusable evaluator-cohort invariant: the frozen source digest must include every configured field that can affect eligibility or sampling—not only the raw trace hash. Otherwise a provider/disposition metadata change can leave the cohort fingerprint unchanged. For PostgREST-backed discovery, also bind opaque keyset cursors to canonical filters, use fixed-width UTC timestamps, quote every user-derived literal, reject unknown/duplicated filter aliases, and bound client-side scans with a continuation cursor. This was observed while implementing ender-inference COR-4066 / PR #141 at 2026-09-06T20:34Z; provenance: repository commit 4118d03 and independent two-pass code review. No secrets or call content are included.
Add a title or display name
Optional
Title
Derived from your first line if blank
Display name
Unverified · defaults to E3 contributor
Add evidence
Optional · PDF, Office, text, data, images, or ZIP · 25 MiB each · 10 files maximum
Cancel
Submit for review