frozen_now fixture is unused by every test

[ThatMattCat]

Severity: Low · Area: Tests / quality · Category: dead-code
Location: services/agent/tests/conftest.py:7

What's wrong

The only fixture in conftest.py, frozen_now, is not referenced by any of the 380 tests (tests needing deterministic time construct their own datetimes or monkeypatch _now). Dead scaffolding that suggests shared-time discipline exists when it does not.

How it fails

A contributor assumes time-freezing is centrally handled via this fixture, mixes it with the ad-hoc monkeypatch.setattr(memory_review, "_now", ...) pattern used elsewhere, and gets inconsistent clocks within one test.

Suggested fix

Delete the fixture, or actually adopt it in the memory_review/quiet_hours tests that hand-roll fixed datetimes.

Adversarial verification — both skeptics confirmed

Skeptic 1 (confirmed) — Confirmed. frozen_now is defined at services/agent/tests/conftest.py:7-11 and a repo-wide grep finds zero references outside its own definition — no test takes it as a parameter and no usefixtures markers exist. Meanwhile test_memory_review_handler.py (lines 52, 131, 164) hand-rolls the same frozen instant (2026-04-13 UTC) via monkeypa…

Skeptic 2 (confirmed) — Confirmed: frozen_now in services/agent/tests/conftest.py (lines 7-11) is the file's only fixture and a repo-wide grep finds no test among the 380 test functions that references it — the only hit is its own definition. The ad-hoc pattern the finding describes also exists as claimed: test_memory_review_handler.py lines 52/131/164 monkey…

---

_{Filed from a multi-agent audit of 855f5cc: 16 reviewers over ~42k lines produced 170 raw findings; each was handed to 2 independent agents prompted to refute it, and only findings both confirmed were kept (19 refuted, 10 split-verdict, 127 unique confirmed). Line numbers are 1-indexed against 855f5cc and will drift as the code changes.}