SAFETY-48: the slew step ceiling reaches the integrator, and what it bounds is the LEAD #93

Merged

admin merged 2 commits from SAFETY-48 into release/1.0

2026-09-19 11:05:26 +02:00

Author	SHA1	Message	Date
glenn schrooyenandClaude Opus 5	a6af332048	SAFETY-48: what the lead cap actually does to a frozen integrator, and what the peak figure depends on checks / unit (pull_request) Successful in 7m55s Details Four review findings, none of them in executable code. The control law, the tests and the instruments' arithmetic are untouched; every number below was re-measured against the committed 16-day capture before it was written down. R1 - app/control.py carried a justification the decision record and MUTATION.md habit 22 had already retracted, and it was false about the code it sits in. It said an unconditional placement "would also pull the integrator toward a command the SATURATION freeze is pinning", priced at 118 of 256 250 ticks. The 118 was measured against a different pair; on the pair the sentence names the two forms are bit-identical - 0 of 256 246 ticks, re-measured. And the narrow form that shipped DOES move a saturation-frozen integrator: it runs before the freeze block and is not gated on `frozen`. Of 79 160 frozen ticks the cap changes `i_w` on 1095, and on 679 of those it moves it further out on the side it was already on - putting back magnitude the freeze had just let fall, which the freeze's own rule forbids. Worst case 2424 W, at prev_w -3400 W, grid +4697 W, actual -4144 W, sat_count 2 -> 3: the freeze leaves the integrator at -176 W and the cap drags it to -2600 W. It never overrides a same-cycle refusal (the freeze clipped on 75 512 ticks, the slew bound on none of them), and that is stated too. It is also where SAFETY-13's settle values went: gating the cap on `not frozen` puts test_control.py part 5 back at -1190 W exactly, while part 4(b) does not come back - that one moved for the winding reason. R3 - the record read its own sensitivity table backwards. `fit_plant()` picks lag 7.5 s on battery-power RMS, which separates it from 2.5 s by 0.3 %, while the tool's own printed fidelity measure (export against the recorded 29.43 kWh) prefers 2.5 s by the whole 35 % gap. Swept properly: at every lag <= 5.0 s release/1.0's worst quarter is ALREADY 3434 W and this card changes it by exactly zero. The 3739 -> 3434 W headline exists at one plant. New section 3a carries the table; the export reduction now leads every summary and the peak is presented as lag-dependent. The form-A rejection is strengthened rather than weakened: form A is worse on the billed quarter at every lag where release/1.0 produces a peak at all, and at every top-N cut of the quarters. R2 - DOCS.md stated modelled watts to the owner as measurement. It now leads with the export reduction, says the figures are simulated, and says plainly that no capacity-tariff improvement is promised. R4 - tools/slew_sim.py --slew narrated "the OVER-ABSORPTION does not move, and at the slowest slews it is slightly WORSE" while printing a monotonically falling column. That sentence is TRUE on release/1.0 (reproduced: 958/1107/1133/1058/987/1029/1150/1143/1081) and false under this branch's law (777/734/663/563/553/507/435/358/321), so SAFETY-39 is not retracted: its record gets a forward reference, and the script now reads SLEW_LEAD_STEPS out of the law it imported and narrates from the column it just printed. Both branches of that narration were driven. R5 - three counts in the L table were decoration and are re-driven in the CI image on a `cp -a`, purged, `:ro`, one file at a time: L5 is 5+5 with the function names exchanged and 5+4 with the branch bodies exchanged (two mutations wearing one name, now listed separately), L6 is 6+3. The separable- state count is grid-dependent and three sweeps disagree, so all three are quoted with their grids and the load-bearing claim is WHERE they separate. R7 - "a large import answered too weakly" is not what the aggregate shows. Form A cuts import by 5.45 kWh and export by 5.18 kWh, and paired per tick delivers 8.56 kWh less offtake energy on the ticks release/1.0 was importing. The harm is a tail redistribution inside quarters that already net to offtake. The conclusion stands; the sentence is narrowed to it. Verified: the full suite in the CI image, purged tree, `:ro` mount, one file at a time, exactly the command in .gitea/workflows/checks.yml - 30/30, exit 0. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BwRLw7eKYj2byVdwB6kbjU	2026-09-19 02:37:01 +02:00
glenn schrooyenandClaude Opus 5	03ef61fa3b	SAFETY-48: the slew step ceiling reaches the integrator, and what it bounds is the LEAD `compute()` freezes the integrator with an asymmetric rule and the comment beside it states the requirement outright - "the requirement is that the INTEGRATOR stop accumulating, not only the command" - but `frozen` is computed from ONE signal, `abs(prev_w - actual_w) > saturation_w`, and slew-limiting never enters it. The integrator moves `gain * error` a cycle and the command moves at most `slew_w`, so the integrator outruns the command whenever `\|error\| > slew_w / gain`: 667 W at the live shape, against a measured median grid error of 1114 W at the onset of a recorded episode. SAFETY-39's retune made it MORE frequent, not less - a smaller step is a lower threshold. ⚠️ THE FORM THE CARD ASKED FOR RAISES THE BILLED QUANTITY, WHICH IS WHY IT IS NOT WHAT SHIPPED. Reusing the asymmetric freeze on a carried slew flag was built from release/1.0's own source and measured over 16.00 days / 256 252 meter instants of the live install's recorder: export 39.61 -> 34.43 kWh, and the worst quarter-hour average OFFTAKE 3739 -> 3935 W with the mean quarter +2.73 W, in five of six plant and tuning variants. A frozen integrator stops at `prev + gain*error`, which undershoots by (1 - gain); for an export that is free and for an import it is offtake, for as long as the ramp lasts. The card said "do not invent a second rule"; the measurement says to, and both sentences are recorded. WHAT SHIPS BOUNDS THE LEAD AT ONE SLEW STEP, inside the branch where the slew actually bound. Same 16 days, same ticks, same disturbance, same plant, same seed: export 39.61 -> 34.07 kWh, wasted throughput 4.67 -> 3.19 kWh, battery sign reversals 2992 -> 2770, median recovery 232 -> 199 s, p99 849 -> 850 s, max 901 -> 898 s, unrecovered 53 -> 31, worst \|i_w - prev_w\| 9200 -> 800 W - and the worst 15-minute average offtake 3739 -> 3434 W, i.e. the billed quantity IMPROVES. `i_w = slewed` (a cap of zero) is better again at 32.02 kWh and is refused: it is the velocity form this module's docstring says `i_w=None` collapses to. TWO COMMITTED NUMBERS MOVED, DELIBERATELY, WITH THE REASON BESIDE THEM. SAFETY-13's released command settles at -800 W and -1200 W where release/1.0 settled at -1190 W and -2390 W, because there is less wound integrator to settle onto; AC 4's property - it stops at the frozen integrator, not the rail - is what is asserted and the second is now graded against `i_w` rather than a literal. And SAFETY-39's "the slew does not own the over-absorption" is now FALSE - INVERTED, on the independent first-order plant test_control.py writes itself: 1005/1109/1285/1330 W rising with a slower slew becomes 729/550/424/326 falling. Two plant models that share nothing agree on the sign. Evidence: `tools/replay_recorded_loop.py` plus the read-only capture it made (scenarios/loop-2026-09-02_2026-09-17). The "before" law is the shipped file with the block removed by a named substitution - never a transcription - and a missing anchor is a REFUSAL carrying "NOTHING HAS BEEN MEASURED". The instrument prints its own fidelity first: it reproduces 39.6 kWh of export where the install exported 29.4, i.e. it is ~35 % pessimistic about the loop's tracking, so only the delta is claimed. 30/30 suites green, including test_control.py's exhaustive i_w=None equivalence sweep against the transcribed release/1.0 law, untouched: the cap moves `i_w` and not `target`, and that sweep compares (target_w, sat_count, frozen, reason). Six law mutations, one instrument battery. L4 - the cap applied unconditionally, -92 B - was a SURVIVOR of all thirty files: it is bit-identical on the shipped configuration (0 of 256 246 ticks) and separable on 2050 of 257 400 swept states, all at `integrator_max_w` above `max_w`, which part 4 now drives. L5 and L6 are ZERO BYTES. AND TWO COMMENTS IN THIS CARD WERE INVENTED AND ARE CORRECTED IN PLACE. The one-sided spelling was justified by a two-sided form that would "drag a trailing integrator" - it does not, and the two agree on 205 920 of 205 920 swept states; and a "118 of 256 252 ticks" figure had been measured against a different pair. MUTATION.md gains habits 21 and 22 for both shapes. NOT DEPLOYED, NOT VERSIONED. The live install still runs release/1.0's law. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BwRLw7eKYj2byVdwB6kbjU	2026-09-19 02:36:28 +02:00

Author

SHA1

Message

Date

glenn schrooyenandClaude Opus 5

a6af332048

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

checks / unit (pull_request) Successful in 7m55s

Details

Four review findings, none of them in executable code. The control law, the
tests and the instruments' arithmetic are untouched; every number below was
re-measured against the committed 16-day capture before it was written down.

R1 - app/control.py carried a justification the decision record and MUTATION.md
habit 22 had already retracted, and it was false about the code it sits in. It
said an unconditional placement "would also pull the integrator toward a
command the SATURATION freeze is pinning", priced at 118 of 256 250 ticks. The
118 was measured against a different pair; on the pair the sentence names the
two forms are bit-identical - 0 of 256 246 ticks, re-measured. And the narrow
form that shipped DOES move a saturation-frozen integrator: it runs before the
freeze block and is not gated on `frozen`. Of 79 160 frozen ticks the cap
changes `i_w` on 1095, and on 679 of those it moves it further out on the side
it was already on - putting back magnitude the freeze had just let fall, which
the freeze's own rule forbids. Worst case 2424 W, at prev_w -3400 W, grid
+4697 W, actual -4144 W, sat_count 2 -> 3: the freeze leaves the integrator at
-176 W and the cap drags it to -2600 W. It never overrides a same-cycle refusal
(the freeze clipped on 75 512 ticks, the slew bound on none of them), and that
is stated too. It is also where SAFETY-13's settle values went: gating the cap
on `not frozen` puts test_control.py part 5 back at -1190 W exactly, while part
4(b) does not come back - that one moved for the winding reason.

R3 - the record read its own sensitivity table backwards. `fit_plant()` picks
lag 7.5 s on battery-power RMS, which separates it from 2.5 s by 0.3 %, while
the tool's own printed fidelity measure (export against the recorded 29.43 kWh)
prefers 2.5 s by the whole 35 % gap. Swept properly: at every lag <= 5.0 s
release/1.0's worst quarter is ALREADY 3434 W and this card changes it by
exactly zero. The 3739 -> 3434 W headline exists at one plant. New section 3a
carries the table; the export reduction now leads every summary and the peak is
presented as lag-dependent. The form-A rejection is strengthened rather than
weakened: form A is worse on the billed quarter at every lag where release/1.0
produces a peak at all, and at every top-N cut of the quarters.

R2 - DOCS.md stated modelled watts to the owner as measurement. It now leads
with the export reduction, says the figures are simulated, and says plainly
that no capacity-tariff improvement is promised.

R4 - tools/slew_sim.py --slew narrated "the OVER-ABSORPTION does not move, and
at the slowest slews it is slightly WORSE" while printing a monotonically
falling column. That sentence is TRUE on release/1.0 (reproduced:
958/1107/1133/1058/987/1029/1150/1143/1081) and false under this branch's law
(777/734/663/563/553/507/435/358/321), so SAFETY-39 is not retracted: its
record gets a forward reference, and the script now reads SLEW_LEAD_STEPS out
of the law it imported and narrates from the column it just printed. Both
branches of that narration were driven.

R5 - three counts in the L table were decoration and are re-driven in the CI
image on a `cp -a`, purged, `:ro`, one file at a time: L5 is 5+5 with the
function names exchanged and 5+4 with the branch bodies exchanged (two
mutations wearing one name, now listed separately), L6 is 6+3. The separable-
state count is grid-dependent and three sweeps disagree, so all three are
quoted with their grids and the load-bearing claim is WHERE they separate.

R7 - "a large import answered too weakly" is not what the aggregate shows. Form
A cuts import by 5.45 kWh and export by 5.18 kWh, and paired per tick delivers
8.56 kWh less offtake energy on the ticks release/1.0 was importing. The harm
is a tail redistribution inside quarters that already net to offtake. The
conclusion stands; the sentence is narrowed to it.

Verified: the full suite in the CI image, purged tree, `:ro` mount, one file at
a time, exactly the command in .gitea/workflows/checks.yml - 30/30, exit 0.

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

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

glenn schrooyenandClaude Opus 5

03ef61fa3b

SAFETY-48: the slew step ceiling reaches the integrator, and what it bounds is the LEAD

`compute()` freezes the integrator with an asymmetric rule and the comment
beside it states the requirement outright - "the requirement is that the
INTEGRATOR stop accumulating, not only the command" - but `frozen` is computed
from ONE signal, `abs(prev_w - actual_w) > saturation_w`, and slew-limiting
never enters it. The integrator moves `gain * error` a cycle and the command
moves at most `slew_w`, so the integrator outruns the command whenever
`|error| > slew_w / gain`: 667 W at the live shape, against a measured median
grid error of 1114 W at the onset of a recorded episode. SAFETY-39's retune
made it MORE frequent, not less - a smaller step is a lower threshold.

⚠️ THE FORM THE CARD ASKED FOR RAISES THE BILLED QUANTITY, WHICH IS WHY IT IS
NOT WHAT SHIPPED. Reusing the asymmetric freeze on a carried slew flag was
built from release/1.0's own source and measured over 16.00 days / 256 252
meter instants of the live install's recorder: export 39.61 -> 34.43 kWh, and
the worst quarter-hour average OFFTAKE 3739 -> 3935 W with the mean quarter
+2.73 W, in five of six plant and tuning variants. A frozen integrator stops at
`prev + gain*error`, which undershoots by (1 - gain); for an export that is
free and for an import it is offtake, for as long as the ramp lasts. The card
said "do not invent a second rule"; the measurement says to, and both sentences
are recorded.

WHAT SHIPS BOUNDS THE LEAD AT ONE SLEW STEP, inside the branch where the slew
actually bound. Same 16 days, same ticks, same disturbance, same plant, same
seed: export 39.61 -> 34.07 kWh, wasted throughput 4.67 -> 3.19 kWh, battery
sign reversals 2992 -> 2770, median recovery 232 -> 199 s, p99 849 -> 850 s,
max 901 -> 898 s, unrecovered 53 -> 31, worst |i_w - prev_w| 9200 -> 800 W -
and the worst 15-minute average offtake 3739 -> 3434 W, i.e. the billed
quantity IMPROVES. `i_w = slewed` (a cap of zero) is better again at 32.02 kWh
and is refused: it is the velocity form this module's docstring says `i_w=None`
collapses to.

TWO COMMITTED NUMBERS MOVED, DELIBERATELY, WITH THE REASON BESIDE THEM.
SAFETY-13's released command settles at -800 W and -1200 W where release/1.0
settled at -1190 W and -2390 W, because there is less wound integrator to
settle onto; AC 4's property - it stops at the frozen integrator, not the rail
- is what is asserted and the second is now graded against `i_w` rather than a
literal. And SAFETY-39's "the slew does not own the over-absorption" is now
FALSE - INVERTED, on the independent first-order plant test_control.py writes
itself: 1005/1109/1285/1330 W rising with a slower slew becomes 729/550/424/326
falling. Two plant models that share nothing agree on the sign.

Evidence: `tools/replay_recorded_loop.py` plus the read-only capture it made
(scenarios/loop-2026-09-02_2026-09-17). The "before" law is the shipped file
with the block removed by a named substitution - never a transcription - and a
missing anchor is a REFUSAL carrying "NOTHING HAS BEEN MEASURED". The
instrument prints its own fidelity first: it reproduces 39.6 kWh of export
where the install exported 29.4, i.e. it is ~35 % pessimistic about the loop's
tracking, so only the delta is claimed.

30/30 suites green, including test_control.py's exhaustive i_w=None
equivalence sweep against the transcribed release/1.0 law, untouched: the cap
moves `i_w` and not `target`, and that sweep compares (target_w, sat_count,
frozen, reason).

Six law mutations, one instrument battery. L4 - the cap applied
unconditionally, -92 B - was a SURVIVOR of all thirty files: it is
bit-identical on the shipped configuration (0 of 256 246 ticks) and separable
on 2050 of 257 400 swept states, all at `integrator_max_w` above `max_w`, which
part 4 now drives. L5 and L6 are ZERO BYTES.

AND TWO COMMENTS IN THIS CARD WERE INVENTED AND ARE CORRECTED IN PLACE. The
one-sided spelling was justified by a two-sided form that would "drag a
trailing integrator" - it does not, and the two agree on 205 920 of 205 920
swept states; and a "118 of 256 252 ticks" figure had been measured against a
different pair. MUTATION.md gains habits 21 and 22 for both shapes.

NOT DEPLOYED, NOT VERSIONED. The live install still runs release/1.0's law.

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

2026-09-19 02:36:28 +02:00