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
{{ message }}
Repository navigation
Commit 194454d
Browse filesBrowse the repository at this point in the historyBrowse files
Phase B/1: floor to >=2.7.1, capability block deleted, resolution re-proved
THE FLOOR IS 2.7.1, NOT 2.7.0. Checked PyPI at execution time as
instructed: the first check showed 2.7.0 (uploaded 17:48:38Z) and no
2.7.1. The clean install minutes later resolved 2.7.1 — the fast-follow
landed at 18:24:03Z, between the check and the install. Re-verified, and
floored on the later one per the rule: the fleet's floor round happens
once.
2.7.0 is what this app cannot START without — configure_geo, the operator
panel, per-vendor policy and the rate ceiling are all called
unconditionally. 2.7.1 is additive (rel="describedby" discovery relations
on both document lanes plus Link headers, a text/plain Accept ramp, and a
source digest that makes representation parity provable). Nothing in this
app needed changing for it; the suite proves that.
Moved in EVERY encoding, found by grepping the number rather than by
memory — which turned up a third, staler one nobody had touched:
requirements.txt >=2.6.1 -> >=2.7.1 (all four extras)
run.py LLMS_PKG_FLOOR (2,6,1) -> (2,7,1), message rewritten
ci.yml the install pin AND both version asserts, which were
still on >=2.3.4 from the template
THE CAPABILITY BLOCK IS GONE. That was its design promise, not a
convenience — a guard kept past the condition it guards is a guard nobody
can reason about. Collapsed to a plain import, and every guarded call site
with it:
- run.py's try/except + LLMS_HAS_27 + _llms_config_27() kwarg filter;
- the three vendor_policy / effective_policies fallbacks in the
showcase exec modules, which existed only to keep the app booting on
the old floor;
- tests/conftest.py's requires_dimll_27 skipif and the pytestmark in
six test modules. That marker could no longer fire — the floor makes
it unreachable — and a skipif that can never trigger is a test suite
quietly lying about its own coverage.
THE RE-PROOF. Phase 1 mirrored the boilerplate's resolved venv because
the sandbox had no network, and said so. This is the real thing: fresh
venv on Python 3.12 (CI's primary, not phase 1's 3.11), full resolve from
PyPI. It pulled genuinely newer transitives than the mirror — DMC 2.8.0,
plotly 6.9.0, pandas 3.0.5, numpy 2.5.2, gunicorn 26.1.0 — and the suite
is green on all of it.
597 passed / 1 skipped (flask)
594 passed / 4 skipped (fastapi, ASGI extras installed as ci.yml does)
audit_links: 0 broken internal, 0 external, 0 anchors
Also verified, since the soak's attestation was on a local file: PyPI's
2.7.0 wheel has a different sha256 from the one soaked
(e9c69d18… vs 05180075…) but the SAME size, and extracting it shows all
25 source files byte-identical. A rebuild at upload time, not different
code.
0 commit comments