Version 1.1.0 in pyproject, the package, uv.lock and every README install
line; the "current tag" entry moves to v1.1.0 and v1.0.0 becomes history.
The CHANGELOG states what a user notices: BM25 ranking on passages and
titles, sub-questions in one call, a map per bundle, a weak-coverage
signal, one generic skill plus the server as the standard way in over a
folder, --gate in okf project, and okf-consumption/2 as BREAKING for a
reader of withheld. Known limitations gain the vocabulary route through
the map, the retrieval gate's red rows as they stand under BM25, the
answer smoke test (no corpus content, no counts) and unmeasured speed.
eval/svar/ is ignored the way eval/soek's sets are: answers about a
consumer's corpus are inputs, never committed. No tag, no push.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`eval/soek/` is where the search gate's question sets are placed by hand. The
sets themselves are `.gitignore`d rather than committed, and that is this
repository's existing rule rather than a new one: `okf_retrieval_gate.py`
already states that a gold set names documents in someone's corpus, so it
arrives as a path and is never a constant here.
Measured before it was decided: the sets carry coordination ids and an
absolute path under a home directory, both of which this repository's own
frames forbid in a tracked file. A set that is absent is reported
`IKKE KJOERT -- fixture mangler` and counted RED, never as 0 hits and never as
a quiet pass, so a machine without them cannot read silence as a green row.
The README states the SHAPE each set must have and nothing about where any of
them came from or what any of them contains.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
