You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The public suite passes in the public repository. 0.4.0's cut shipped five tests that
load a script (scripts/mint_archetype_sandbox.py, which carries a live address) or an
asset (the portal's built sprite) the cut does not carry, and the public repository's
first CI run failed on collection. Those tests are FND's (the ratchet now says so: a
test that reaches a script or a path the cut does not ship is not public), and scripts/publish_micyte.py runs micyte/tests inside the export before it builds the
wheel, so the release gate refuses a cut whose suite would not pass where it is
published.
The class library and the viewscopes keep their families. 0.4.0's readdress moved 63 kind_* and viewscope_* documents with the archetypes, and their readers — which
select rows BY FAMILY (header 4-1, members 4-2, slots 4-3; container 4-1, slots 4-2),
as their modules say — went blind: every viewscope pane drew zero groups and the
calendar found zero logs, while the archetype registry parsed identically. The library
was restored from the pre-window backup and readdressed again under one rule, datum_ops.positional_grammars: a document a positional reader owns is not held to I7.
The readdress refuses such a document whatever its flags say, the audit reports it
apart instead of as a finding, and the lock is re-cut — the 57 archetypes keep 0.4.0's
hashes (the packages' pins are unchanged), the 63 owned documents return to theirs.
No engine change.