# Retired claims. One term per line; blank lines and #-comments ignored.
# Matching is case-insensitive substring, against README.md, setup.py, and
# pyproject.toml's [project].description + .keywords (the fields that become the
# PyPI project page).
#
# A term earns a place here when it was withdrawn from the portfolio for being
# wrong or unsupportable -- not merely for being out of fashion. The point is
# that a retired claim tends to survive in metadata long after the README that
# carried it was corrected, because nobody re-reads a `keywords` array.
#
# All three seeds below were live in the portfolio as of the August 2 audit,
# each of them sitting above an already-corrected README:
#   CET1                                 -- cdfi-benchmark's PyPI description.
#                                           CET1 is a Basel III bank-capital
#                                           ratio; the package reports Tier 1
#                                           leverage off FDIC call reports and
#                                           never computed CET1.
#   disparate-impact                     -- fair-lending-screener keyword. The
#                                           package computes descriptive
#                                           distributions; disparate impact is a
#                                           legal finding it does not make.
#   the methodology federal examiners use -- fair-lending-screener's GitHub
#                                           description. No examiner methodology
#                                           was implemented or verified.
#
# nmtc-mapper carries none of these today (verified). They are seeded here so
# the gate travels with the meaning intact when this file is copied to the next
# repo, where they ARE live.
CET1
disparate-impact
the methodology federal examiners use
