glenn schrooyenandClaude Opus 5 9099ceb1af SAFETY-31: the grid is 42, not 90, and the wrong number was in the failure text ...

Review reproduced every figure in both PRs except one, and rejected on it.

⚠️ "0 of 90 GAINS x SPLITS combinations produce opposing shares" - the SUBSTANCE
is right and review re-measured it, but the DENOMINATOR is invented. sitesim.py
declares GAINS with 7 entries and SPLITS with 6: the grid is 42. 90 is not even
divisible by 7, so no reading of "GAINS x SPLITS" reaches it. It is almost
certainly TEST-20`s RUN count (84 + 6), which is a different quantity that
happened to be nearby.

⚠️ AND THE WORST COPY OF IT WAS INSIDE THE `PRECONDITION FAILED` TEXT, so the
next engineer to see this scenario go red would have been handed an
unreproducible measurement as their diagnosis - the exact defect class this
campaign exists to correct, in the failure path where it does the most harm.

Corrected in all seven places. At the failure-text site the denominator is now
`len(GAINS) * len(SPLITS)`, computed rather than typed, so it cannot drift from
the grid again - the same fix test_translations.py already took.

Also N3: a missing blank line was going to render the S3 and M19 paragraphs as
one in DECISIONS.

Re-run after the change: `inject.py two_master_circulation` PASS in 35 s, norig,
no rig lock taken. Add-on suite 26 files, all pass.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BwRLw7eKYj2byVdwB6kbjU

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

..

app

OPT-09: the 6500 W window is a bound the ratings never let the site reach

2026-09-06 23:26:22 +02:00

DECISIONS

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

scenarios

DATA-01: thirty days of the real meter, and what AC 3 actually measures

2026-08-31 11:15:02 +02:00

tools

PKG-13: one safe.directory flag, not two, and the tighter of the pair

2026-09-07 17:27:36 +02:00

translations

OPT-09: the weights parameter nothing ever passed, and what it cannot do

2026-09-06 22:55:20 +02:00

build.yaml

Packaged as a Home Assistant add-on, with a field guide

2026-08-23 01:15:25 +02:00

CHANGELOG.md

PKG-12: the live options WERE checked, and the changelog said they were not

2026-09-07 00:47:11 +02:00

config.yaml

PKG-12: bump 0.8.0 -> 0.9.0, so the five merges behind it can actually ship

2026-09-07 00:18:36 +02:00

Dockerfile

Six silent failures found by installing this on a live system

2026-08-23 02:25:58 +02:00

DOCS.md

PKG-09: refuse a deploy that would be a no-op, or that the add-on would reject

2026-09-06 23:57:51 +02:00

MUTATION.md

SAFETY-37: a stale reading is None, and arrival cannot be what says so

2026-09-06 11:53:26 +02:00

requirements.txt

Six silent failures found by installing this on a live system

2026-08-23 02:25:58 +02:00

run.sh

Six silent failures found by installing this on a live system

2026-08-23 02:25:58 +02:00

test_alloc.py

SAFETY-39: slew_w is a ratio to the rail, and gain is what is still wrong

2026-09-06 01:10:15 +02:00

test_arbiter.py

OPT-09: the weights parameter nothing ever passed, and what it cannot do

2026-09-06 22:55:20 +02:00

test_arbitrage.py

TEST-16: pin the two one-sided arbitrage rungs, and un-mask them from each other

2026-09-01 07:37:19 +02:00

test_bytecode.py

TEST-15: purge the bytecode cache before a mutation run, not after

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

test_ci_gate.py

CI-02: grade the merge, and stop a cancelled run reading as a test failure

2026-09-01 03:42:54 +02:00

test_clamp.py

SAFETY-38 round 4: the enumeration covered one sink, and it was not the one this codebase uses

2026-09-05 22:38:40 +02:00

test_control.py

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

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

test_delivery.py

OPT-09: the weights parameter nothing ever passed, and what it cannot do

2026-09-06 22:55:20 +02:00

test_freeze.py

SAFETY-39: slew_w is a ratio to the rail, and gain is what is still wrong

2026-09-06 01:10:15 +02:00

test_freshness.py

SAFETY-37: a stale reading is None, and arrival cannot be what says so

2026-09-06 11:53:26 +02:00

test_inverter.py

OPT-09: the weights parameter nothing ever passed, and what it cannot do

2026-09-06 22:55:20 +02:00

test_loop.py

SAFETY-39: slew_w is a ratio to the rail, and gain is what is still wrong

2026-09-06 01:10:15 +02:00

test_maintenance.py

SPIKE-04a: inverter B is 4600 W FROM ITS DATASHEET, and the copies are pinned

2026-09-04 22:19:31 +02:00

test_mqtt.py

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

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

test_p1.py

TEST-15: purge the bytecode cache before a mutation run, not after

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

test_packshare.py

OPT-09: the 6500 W window is a bound the ratings never let the site reach

2026-09-06 23:26:22 +02:00

test_peakguard.py

OPT-09: the weights parameter nothing ever passed, and what it cannot do

2026-09-06 22:55:20 +02:00

test_price_forecast.py

TEST-15: purge the bytecode cache before a mutation run, not after

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

test_price.py

TEST-15: purge the bytecode cache before a mutation run, not after

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

test_selfconfig.py

UX-02: pin both halves of the max_w fix, because neither was graded

2026-09-06 22:38:54 +02:00

test_static_tariff.py

TEST-16: pin the two one-sided arbitrage rungs, and un-mask them from each other

2026-09-01 07:37:19 +02:00

test_tariff.py

CFG-04: an operator can declare a second inverter, and the shape was measured

2026-09-04 04:29:45 +02:00

test_translations.py

OPT-09: the weights parameter nothing ever passed, and what it cannot do

2026-09-06 22:55:20 +02:00

test_validate.py

OPT-09: the weights parameter nothing ever passed, and what it cannot do

2026-09-06 22:55:20 +02:00

test_version_gate.py

PKG-13: the no-op half of the pre-deploy check has never run as documented

2026-09-07 17:02:01 +02:00

test_watchdog.py

SAFETY-31 AC 4: the alarm's attribution was graded by nothing, and a mutation proved it

2026-09-07 19:08:33 +02:00