glenn schrooyenandClaude Opus 5.5 4707d37019 SAFETY-95 review: test the stale-link and contradiction legs; refusal +1 only ...

F1: test_safety95 checks that a dropped draining battery is this rule's
(stale link, N=3, 4000 W: grid import 676 W vs base 786 W); kills R2.
F2: _charge_track tested directly with a contradicting drain (SAFETY-91's
pin as a floor against a closed pack); kills R3. Today a drain contradicts
only through a refusal; the clause stays, as a contradiction sends 0 W.
F3: a drain ends on a refusal only when the battery will not give
(refused > 0, as _held_back()); one that will not take still drains.
Tested both signs; kills R8.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

2026-10-07 16:41:16 +02:00

..

CFG-04-inverter-roster.md

CFG-04 fix round: the probe now measures the key it is cited for

2026-09-04 10:33:51 +02:00

FCST-04-the-day-that-is-carried.md

TARIFF-07: three plugins that never fetched a price are withdrawn, and a site configured on one is told so

2026-09-13 02:46:15 +02:00

FCST-06-cannot-persist-is-one-question.md

FCST-06: "cannot persist" becomes one question, asked once, about the outcome

2026-09-10 08:43:17 +02:00

MAINT-05-per-pack-maintenance.md

SAFETY-78: maintenance drain at N=2 no longer stalls

2026-09-29 19:52:51 +02:00

MAINT-10-local-start-hour.md

MAINT-10 review: warn when HA did not answer and TZ is UTC; test the 3 s budget

2026-10-05 23:24:02 +02:00

OPT-09-soc-aware-weights.md

SAFETY-43: delinquency measured against the ask, not against a flat 500 W

2026-09-08 03:14:16 +02:00

OPT-11-battery-modes.md

OPT-11 P4 review: records and text

2026-09-29 22:29:00 +02:00

SAFETY-10-shadow-mode.md

SAFETY-47: what a shadow run can show, and the watts on its status row

2026-09-15 01:26:57 +02:00

SAFETY-13-bounding-the-output-freeze.md

SAFETY-13: the bound came from the wrong half of a two-sided measurement

2026-09-06 18:32:03 +02:00

SAFETY-25-loop-soc-floor.md

SAFETY-25: record the owner's decision - option a, with alloc-01 as the revisit trigger

2026-09-02 00:46:54 +02:00

SAFETY-30-per-unit-sign.md

SAFETY-30: pin WHICH thresholds each sign detector is armed with, not just what the numbers are

2026-09-04 11:13:13 +02:00

SAFETY-31-circulating-power.md

SAFETY-31: the grid is 42, not 90, and the wrong number was in the failure text

2026-09-07 20:13:01 +02:00

SAFETY-32-per-pack-bms-floor.md

SAFETY-87: peak guard uses every battery that can still give

2026-10-01 09:29:30 +02:00

SAFETY-37-freshness-policy.md

SAFETY-37: correct a wall-clock tick and a false size claim

2026-09-06 12:35:27 +02:00

SAFETY-38-site-standing-pin.md

SAFETY-38: correct the survivor count, which was stated four ways

2026-09-05 23:10:30 +02:00

SAFETY-39-slew-against-the-rail.md

SAFETY-48: what the lead cap actually does to a frozen integrator, and what the peak figure depends on

2026-09-19 02:37:01 +02:00

SAFETY-40-the-integrator-bound-is-not-a-bug.md

SAFETY-40 review: integrator-bound WARNING gets its own notification id

2026-09-26 23:08:44 +02:00

SAFETY-42-unit-delivery.md

SAFETY-43: delinquency measured against the ask, not against a flat 500 W

2026-09-08 03:14:16 +02:00

SAFETY-43-delinquency-relative-to-the-ask.md

SAFETY-43 round 2: the cut-out "positive control" was vacuous, and the record

2026-09-08 04:15:07 +02:00

SAFETY-45-sizing-the-trade.md

SAFETY-45: a price-driven charge is SIZED by the capacity guard, not switched by it

2026-09-14 16:10:34 +02:00

SAFETY-47-what-a-shadow-run-can-show.md

SAFETY-47: what a shadow run can show, and the watts on its status row

2026-09-15 01:26:57 +02:00

SAFETY-48-bounding-the-integrator-lead.md

SAFETY-48: what the lead cap actually does to a frozen integrator, and what the peak figure depends on

2026-09-19 02:37:01 +02:00

SAFETY-52-a-battery-that-refuses.md

SAFETY-52 review: a refusal is under 100 W, and any delivery ends it

2026-09-27 18:11:42 +02:00

SAFETY-56-what-the-gate-cannot-tell-apart.md

SAFETY-56: the gate cannot tell a dead register from a pack that cannot comply, so stop saying it can

2026-09-23 21:02:40 +02:00

SAFETY-57-the-exit-from-a-stale-verdict.md

SAFETY-56: the gate cannot tell a dead register from a pack that cannot comply, so stop saying it can

2026-09-23 21:02:40 +02:00

SAFETY-61-control-task-death.md

SAFETY-61: the control task's death ends the process, after the wind-down

2026-09-24 23:52:42 +02:00

SAFETY-73-refusal-per-inverter.md

SAFETY-73 review: a refusal needs 22 s of evidence, and the cut-out warning is for giving only

2026-09-28 02:16:24 +02:00

SAFETY-78-drain-at-two.md

SAFETY-84: maintenance charge at N=2 no longer stalls

2026-09-30 19:47:15 +02:00

SAFETY-79-peak-margin.md

SAFETY-79 review: margin under a banked month peak too; identity harness reaches headroom and maintenance

2026-09-29 00:49:16 +02:00

SAFETY-81-load-too-high.md

SAFETY-81 review: 10 s hold, ERROR line per episode, more scenarios

2026-09-29 20:58:37 +02:00

SAFETY-83-peak-over-drain.md

SAFETY-86: peak guard outranks the maintenance hold

2026-10-01 00:56:34 +02:00

SAFETY-84-charge-at-two.md

SAFETY-86: peak guard outranks the maintenance hold

2026-10-01 00:56:34 +02:00

SAFETY-85-no-false-not-keeping-up.md

SAFETY-85 review: grade discharge only, over the plant's settle time

2026-10-01 20:25:41 +02:00

SAFETY-86-peak-over-hold.md

SAFETY-86 review: hold the site at the peak floor while a HOLD gives

2026-10-01 02:00:20 +02:00

SAFETY-87-guard-uses-able-batteries.md

SAFETY-87 review: maintenance pins beside a held battery

2026-10-01 10:50:01 +02:00

SAFETY-90-peak-guard-after-a-gap.md

SAFETY-90 review: count the start-up seconds like a gap

2026-10-05 12:24:33 +02:00

SAFETY-91-held-battery-last.md

SAFETY-91 review: a held-back battery is released 1 point above its reserve

2026-10-02 14:54:20 +02:00

SAFETY-93-charge-end.md

SAFETY-93 review: a charge that ends by a stale link or a refusal hands back smoothly too

2026-10-03 00:10:21 +02:00

SAFETY-94-lagging-battery.md

SAFETY-97: "gives nothing" is graded against each battery's idle reading

2026-10-04 12:21:24 +02:00

SAFETY-95-drain-end.md

SAFETY-95 review: test the stale-link and contradiction legs; refusal +1 only

2026-10-07 16:41:16 +02:00

SAFETY-96-dropped-battery.md

SAFETY-96 review: keep the refusal count over a refused write; test a fast board

2026-10-06 05:46:32 +02:00

SAFETY-97-idle-offset.md

SAFETY-97: centre the band on the idle reading clamped to 75 W

2026-10-04 16:57:16 +02:00

TARIFF-07-withdrawing-three-plugins.md

TARIFF-07: three plugins that never fetched a price are withdrawn, and a site configured on one is told so

2026-09-13 02:46:15 +02:00

TARIFF-09-the-range-and-the-override.md

TARIFF-09: ask the day-ahead feed for the horizon it was handed

2026-09-10 10:35:12 +02:00

TARIFF-10-fetching-is-not-trading.md

TARIFF-10: configuring a price provider is what fetches prices

2026-09-12 20:58:48 +02:00

TARIFF-12-a-degradation-is-not-a-refresh.md

Merge release/1.0 (TARIFF-10) into TARIFF-12

2026-09-12 21:31:27 +02:00

TARIFF-13-the-coverage-of-a-peak.md

TARIFF-13: the coverage figure beside the peak stops resetting at a restart

2026-09-13 12:54:15 +02:00

TARIFF-14-the-window-a-point-covers.md

Merge release/1.0 (TARIFF-12) into TARIFF-14

2026-09-13 01:47:17 +02:00

TARIFF-15-the-card-is-the-grader.md

TARIFF-18: price_taxes_eur_kwh stops shipping at half the real figure

2026-09-13 16:24:56 +02:00

TEL-07-second-mqtt-device.md

TEL-07: a second MQTT device for inverter B, and A keeps the one it has

2026-09-05 14:14:53 +02:00

TEL-10-publishing-the-integrator.md

TEL-10: the loop's integrator stops being invisible

2026-09-18 23:36:23 +02:00

TEL-11-pv-production.md

TEL-14: PV sign declared per entity (pv_invert_entities)

2026-09-25 00:25:24 +02:00

TEL-12-several-pv-inverters.md

UX-08 phase 2b review: keep over-rating evidence, pin latch clearing, plainer rows

2026-09-27 03:10:25 +02:00

TEL-14-pv-sign.md

TEL-14 N1: trim DECISIONS (drop probe transcript, §1 to the reason), move the mutation ledger there with M2/M5

2026-09-25 01:11:40 +02:00

UX-02-two-config-surfaces.md

UX-15: setup-only mode, and the wizard fixes F1, F2, meter, step 1

2026-09-28 08:48:21 +02:00

UX-04-the-configuration-tab-and-the-defaults-pin.md

UX-04: nothing was hidden, the defaults pin was never built, and CFG-04 was closed by UX-02

2026-09-07 18:29:48 +02:00

UX-08-short-notifications.md

UX-11 review: restored stops log their figures to events, the quiet-meter WARNING is rate-limited

2026-09-27 14:21:05 +02:00

UX-13-one-layout-table.md

UX-13 P3 review: truthful peak lines, a clock-proof test, Simple stays default

2026-09-28 23:20:33 +02:00

UX-13-setup-wizard.md

UX-15: setup-only mode, and the wizard fixes F1, F2, meter, step 1

2026-09-28 08:48:21 +02:00

UX-15-setup-only-mode.md

UX-15 review: DECISIONS states the TEST-23 run that happened (PR #147)

2026-09-28 15:15:51 +02:00