clay-borg/tools/rule-coverage.py

416 lines
17 KiB
Python
Raw Normal View History

#!/usr/bin/env python3
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
"""AM-1 / M-D1-COV and AM-1b / M-D1-LNK, over **every** numbered spec.
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
Two denominators, deliberately separate because they are measured
differently:
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
ground GR-rules in specs/GroundRules.md, covered by `covers:` lists in
scenarios/ground/*.yaml, and linked to the aggregate source.
kernel K-rules in specs/GameKernel.md. **Link only.** K-rules are
kernel invariants, not game rules: there is no kernel aggregate,
no setup preset and no command vocabulary, so a
`scenarios/kernel/*.yaml` with `covers: [K11]` would be a tag in
a directory the runner cannot dispatch. Claiming scenario
coverage for them would be the inflation this gate exists to
prevent.
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
Until CB-WP-0005 T01, `AGGREGATE` was one file and the rule pattern
matched `GR-` only, so the kernel spec was outside the instrument
entirely. K10, K14 and K18 were unimplemented for four workplans while
`make coverage` printed `58/58 (100%)`.
Stated limit (InnerLoop implementation rule 4), now doubly important:
**this gate counts names.** It proves no rule is unclaimed, no claimed
rule is invented, and no rule is absent from the source. It does NOT
prove anything fails when a rule is violated that is M-D1-MUT
(`make mutation-check`, CB-WP-0005 T02), and four of the seven defects
found in CB-RES-0004 were invisible to a name-based check.
Positive control (InnerLoop v1.1 §Step 5): every denominator asserts it
actually found rules. Before this was added, a broken spec regex yielded
rules=[] and missing=[] and the tool exited 0 reporting "0/0" the
harness-does-nothing class, in the tool that reports our headline
coverage number. A new denominator inherits the control, or it
reintroduces the defect the old one was fixed for.
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
Usage:
python3 tools/rule-coverage.py
python3 tools/rule-coverage.py --self-test
"""
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
import datetime
import glob
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
import os
import re
import sys
2026-07-31 10:13:52 +02:00
from repo import enter_root
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
RULE_RE = r"\*\*(GR-[A-Z]+\d+)"
COVERS_RE = r"covers: \[(.*?)\]"
AGGREGATE = "games/ground/src/lib.rs"
# CB-WP-0003 T08: a provisional default with no expiry can shape the kernel
# indefinitely while looking handled. CI warns; it does not break the build,
# because the ruling is a ground-game decision we cannot make for them.
PROVISIONAL_WARN_DAYS = 30
ID_RE = r"GR-[A-Z]+\d+"
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
# ADR-0005 §5: a newly widened denominator is not a regression, so the
# kernel arm reports without feeding the exit code — but only until a
# date that lives in the tool rather than in prose. An open-ended "we
# will gate it later" is how AM-4's targets went unratified for four
# workplans. The remaining days are printed on every run.
KERNEL_GATES_FROM = datetime.date(2026, 8, 31)
# Source roots searched for rule IDs. A list, not one file: K-rules live
# in cb-kernel, cb-events and cb-game-runtime, so a single-file AGGREGATE
# would report every one of them unlinked forever.
SOURCE_ROOTS = ("crates", "games", "tools")
def source_files(roots=SOURCE_ROOTS):
"""Every .rs file under the given roots, excluding build output."""
out = []
for root in roots:
for dirpath, dirnames, files in os.walk(root):
dirnames[:] = [d for d in dirnames if d != "target"]
out += [os.path.join(dirpath, f) for f in sorted(files)
if f.endswith(".rs")]
return sorted(out)
def parse_rules(spec_text, pattern=RULE_RE):
return sorted(set(re.findall(pattern, spec_text)))
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
def parse_code_ids(text, pattern=ID_RE):
"""Rule IDs named anywhere in the source (T09)."""
return set(re.findall(pattern, text))
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
def code_ids_over(paths, pattern):
"""Union of rule IDs named across many source files."""
found = set()
for p in paths:
found |= parse_code_ids(open(p).read(), pattern)
return found
def provisional_items(paths):
"""(path, owner, raised) for every scenario encoding a U-item default."""
out = []
for path in paths:
text = open(path).read()
if not re.search(r"^provisional:\s*true", text, re.M):
continue
owner = re.search(r"^provisional_owner:\s*(\S+)", text, re.M)
raised = re.search(r"^provisional_raised:\s*(\S+)", text, re.M)
out.append((path,
owner.group(1) if owner else None,
raised.group(1) if raised else None))
return out
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
def parse_covers(text):
match = re.search(COVERS_RE, text, re.S)
if not match:
return set()
return {c.strip() for c in match.group(1).split(",") if c.strip()}
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
def _silent_output():
"""Whatever the kernel arm prints with quiet=True — must be nothing."""
import io
from contextlib import redirect_stdout
buf = io.StringIO()
with redirect_stdout(buf):
kernel_arm(today=datetime.date(2026, 1, 1), quiet=True)
return buf.getvalue()
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
def self_test():
"""Each assertion pins a failure this tool must detect."""
results = []
def check(name, ok, detail=""):
results.append((name, ok, detail))
# The defect that motivated this control: a spec that parses to zero
# rules must not be reportable as coverage.
check("zero rules detected as a failure", parse_rules("no rules here") == [],
"empty spec yields no rules; main() now aborts on this")
# The matcher must actually match the real format.
check("rule matcher works on real spec format",
parse_rules("**GR-R06** something\n**GR-A12** other")
== ["GR-A12", "GR-R06"])
# covers: parsing, including the empty case.
check("covers matcher works", parse_covers("covers: [GR-R06, GR-A12]")
== {"GR-R06", "GR-A12"})
check("missing covers yields empty set", parse_covers("no covers key") == set())
# T09: the spec -> code link must be detectable.
check("code-id matcher finds ids in source",
parse_code_ids("// GR-R06: lead first\nfn f(){} // GR-A12")
== {"GR-R06", "GR-A12"})
check("code-id matcher finds none in unmarked source",
parse_code_ids("fn f() { let x = 1; }") == set())
# T08: every provisional scenario must carry an owner and a date.
CB-WP-0026: collect the rulings -- ten answers that arrived and were never applied ground-game ruled all ten U-items on 2026-08-03, every one CONFIRMED as the default clay-borg simulates, and confirmed five of six provisional scenarios. clay-borg never collected the answers: CB-RES-0007 reported "0 of 10 ruled" the same day, and CB-WP-0022 built the finding register two days later still recording them as `reported`. make design's first run is what noticed -- not a human, not the adversarial review that found four other things. That is the unread-inbox failure running in the opposite direction, and it appears nowhere in the declaration, survey, ADR or spec of the pass that was built entirely around the forward version. It is arguably worse: an unread message is visible as silence, while a collected-but-unapplied ruling looks exactly like work in progress. Ten rulings quoted into §Underdetermined (the three conditional ones verbatim -- U1's designer note, U2's End-only trigger, U8's consume-only-if-it-cancels). Five provisional flags lifted, replaced by ruled/ruled_by/ruled_note so the flag went and the provenance stayed. Register queue 9 -> 0. T02's control came back clean: make sim is 26 passed, 59 rules covered, nothing red. Had a scenario gone red it would have meant we described our own behaviour incorrectly to ground-game. I wrote two U-item mappings and both were wrong. gr-a04 -> U1 (it asserts consent is REQUIRED; U1 asks WHEN the target accepts) and gr-d05 -> U5 (it exercises the UNREJECTED Reverse; U5 is the rejected one). Both plausible from covers:, neither survived reading the description. Third and fourth instance of this defect; the first two reached ground-game. So encodes_u_item is now a declaration and design.py asserts the file names what it claims -- and that check's own first version grepped for mentions and went red when two files recorded why they do NOT encode U1 and U5. A mention is not a claim, which is exactly the looseness that let "six of the ten have provisional scenarios" stand. Two positive controls went red for the best possible reason, both broken the same way -- asserting against live repo data instead of constructing their condition. rule-coverage.py required at least one provisional item to EXIST; it now builds a fixture and reports the live count as a diagnostic, because there is no number of provisional items this project should have. design-baseline.py pinned "2 of 6" while recomputing one row from a live glob, so the dated snapshot was never a snapshot; frozen to its 2026-08-03 list and unwired from self-tests, since per ADR-0012 D8 it is no longer a reporting tool. ScenarioFile is deny_unknown_fields and refused the four new fields until declared -- correct: a corpus accepting unknown metadata would let a typo'd encodes_u_iem sit there claiming nothing. DEVIATION: ADR-0012 D2 said "no new file". GroundRules.md crossed the loadability limit, so the register moved to specs/FindingRegister.md. D2's substance holds -- one register, same machinery, nothing competing -- but the literal instruction did not, and it resolves an awkwardness D2 named itself. make all: exit 0. loop-lint clean. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-05 16:13:37 +02:00
#
# **Tested against a fixture, not against the repo.** This control used
# to read `scenarios/ground/*.yaml` and require `bool(prov)` — a guard
# against a vacuous pass, which is the right instinct, but it made the
# control depend on the repo *having* provisional items. CB-WP-0026
# lifted the last five (ground-game ruled them all on 2026-08-03) and
# the control went red for the best possible reason: nothing is
# provisional any more.
#
# A positive control must construct the condition it tests. The live
# count is a diagnostic below, not a pass criterion — there is no
# number of provisional items this project *should* have.
import glob as _g, tempfile as _tf, os as _os
with _tf.TemporaryDirectory() as _d:
good = _os.path.join(_d, "good.yaml")
bad = _os.path.join(_d, "bad.yaml")
open(good, "w").write("scenario: x\nprovisional: true\n"
"provisional_owner: ground-game\n"
"provisional_raised: 2026-07-31\n")
open(bad, "w").write("scenario: y\nprovisional: true\n")
g = provisional_items([good])
b = provisional_items([bad])
check("a provisional item with an owner and a date passes",
len(g) == 1 and all(o and r for _, o, r in g))
check("a provisional item MISSING its owner/date is caught",
len(b) == 1 and not all(o and r for _, o, r in b),
"without this the check passes on anything it can parse")
live = provisional_items(sorted(_g.glob("scenarios/ground/*.yaml")))
check("every LIVE provisional item has an owner and a raised date",
all(o and r for _, o, r in live),
f"{len(live)} provisional item(s)"
+ (" — none; all ruled 2026-08-03 (CB-WP-0026)" if not live else ""))
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
# --- CB-WP-0005 T01: the kernel denominator ---
# The control the old arm was fixed for, replicated. A pattern that
# stops matching must abort, not report 0/0 as though it were 100%.
check("kernel: zero rules detected as a failure",
parse_rules("no kernel rules here", r"\*\*(K\d+)\*\*") == [])
check("kernel: matcher works on the real spec format",
parse_rules("- **K10** A replay bundle\n- **K9** A snapshot",
r"\*\*(K\d+)\*\*") == ["K10", "K9"])
check("kernel: matcher does not match GR-rules",
parse_rules("**GR-R06** lead first", r"\*\*(K\d+)\*\*") == [])
# A single-file AGGREGATE reported every K-rule unlinked forever; the
# union across roots is the fix, so assert it actually unions.
# Compute once and report the same value that was asserted. Building
# the detail string with a second, re-escaped copy of the pattern
# printed "0 K-ids" beside a passing ">5" assertion — a label that
# contradicts its own check is worse than no label.
srcs = source_files()
k_in_src = code_ids_over(srcs, r"\bK\d+\b")
check("kernel: ids union across many files, not just one",
code_ids_over([os.devnull], r"\bK\d+\b") == set() and len(k_in_src) > 5,
f"{len(k_in_src)} K-ids across {len(srcs)} files")
check("kernel: source roots resolve to real files",
len(srcs) >= 5, f"{len(srcs)} .rs files")
# The gate date must actually change behaviour, in both directions.
# A "binds later" that never binds is the AM-4 failure this replaces.
CB-WP-0006 T07: implement K18, amend K14 Two rules, two different answers, which is the point of a task phrased "implement, or amend and say why". K18 is implemented. "Criterion benches driving the same scenario format at scale" was false — the bench hardcoded its commands and never touched ScenarioFile, while MetricsAndScenarios §3 pointed at a benchmarks/ directory containing only baselines/. benchmarks/synthetic-3p.yaml now holds the workload and both the bench and bench_shape read it: the workload is data, not code. A second defect surfaced while fixing the first. After the bench switched to the file, bench_shape still hardcoded the same sequence, so the workload existed twice — deleting end_round from the YAML broke bench-test while bench_shape kept passing. Duplicated-fact drift in executable form. Both now read the same include_str! and deleting a command breaks both. Explicitly not claimed: this does not unblock AM-3. AM-3's baseline is a declarative game object — moves, turn order, rules. synthetic-3p.yaml is a command list; the rules live in games/ground. Marking it as AM-3's subject would compare a script to a game definition, which is the category error AM-3 is blocked on. The file says so in its own header, where the next person will be tempted. K14 is amended. CommitWindow had zero non-test users and GROUND enforces the same contract inline. Wiring GROUND through it was rejected: it would change the serialized shape of `selections`, which four scenario files assert by dot-path and every state hash depends on, for the sole benefit of making a sentence literally true. The deciding argument is INTENT's, not convenience: abstractions are extracted from working games rather than invented in isolation, and no concept becomes canonical until it survives a second concrete use. CommitWindow was invented before any game needed it and has survived none. Imposing it on GROUND would manufacture the first use rather than discover it. So K14 states what is actually guaranteed, CommitWindow is marked provisional in the source, and it carries a delete-by date of 2026-12-31. Kernel spec->code link 16/18 -> 18/18, stated with the caveat the gate prints every run: that is about names, not assertions. Two self-tests broke and both broke correctly. rule-coverage's gate test hardcoded "unlinked rules exist today" and failed when the last one was linked; it now computes that and asserts the gate fails iff rules are unlinked. facts' text check rejected k_unlinked once it became legitimately empty; empty now renders as "(none)" and the check distinguishes absent from empty. M-D1-MUT: 8 of 14, unchanged — K14 and K18 are kernel rules, not acceptance rows. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-01 12:47:16 +02:00
# What the gate must do depends on whether anything is unlinked, so
# compute that rather than assuming it. An earlier version hardcoded
# "unlinked rules exist today" and failed the moment T07 linked the
# last one — correctly, but for the wrong reason.
_k = parse_rules(open(os.path.join("specs", "GameKernel.md")).read(),
r"\*\*(K\d+)\*\*")
_named = code_ids_over(source_files(), r"\bK\d+\b")
_unlinked = [r for r in _k if r not in _named]
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
before = kernel_arm(today=datetime.date(2026, 1, 1), quiet=True)
after = kernel_arm(today=datetime.date(2027, 1, 1), quiet=True)
# The reporting path must be exercised, not only the quiet one. When
# this control ran `quiet=True` exclusively, a broken `say()` made
# every real `make coverage` die with RecursionError while the
# self-test printed all-ok — a positive control that named the
# behaviour without asserting it, which is exactly the defect
# CB-RES-0004 is about.
import io
from contextlib import redirect_stdout
buf = io.StringIO()
with redirect_stdout(buf):
loud = kernel_arm(today=datetime.date(2026, 1, 1))
out = buf.getvalue()
check("kernel: the reporting path actually prints",
loud == before and "AM-1b kernel spec->code link:" in out
and "gate:" in out,
f"{len(out.splitlines())} lines")
check("kernel: quiet suppresses output, loud does not",
out.strip() != "" and _silent_output() == "")
CB-WP-0006 T07: implement K18, amend K14 Two rules, two different answers, which is the point of a task phrased "implement, or amend and say why". K18 is implemented. "Criterion benches driving the same scenario format at scale" was false — the bench hardcoded its commands and never touched ScenarioFile, while MetricsAndScenarios §3 pointed at a benchmarks/ directory containing only baselines/. benchmarks/synthetic-3p.yaml now holds the workload and both the bench and bench_shape read it: the workload is data, not code. A second defect surfaced while fixing the first. After the bench switched to the file, bench_shape still hardcoded the same sequence, so the workload existed twice — deleting end_round from the YAML broke bench-test while bench_shape kept passing. Duplicated-fact drift in executable form. Both now read the same include_str! and deleting a command breaks both. Explicitly not claimed: this does not unblock AM-3. AM-3's baseline is a declarative game object — moves, turn order, rules. synthetic-3p.yaml is a command list; the rules live in games/ground. Marking it as AM-3's subject would compare a script to a game definition, which is the category error AM-3 is blocked on. The file says so in its own header, where the next person will be tempted. K14 is amended. CommitWindow had zero non-test users and GROUND enforces the same contract inline. Wiring GROUND through it was rejected: it would change the serialized shape of `selections`, which four scenario files assert by dot-path and every state hash depends on, for the sole benefit of making a sentence literally true. The deciding argument is INTENT's, not convenience: abstractions are extracted from working games rather than invented in isolation, and no concept becomes canonical until it survives a second concrete use. CommitWindow was invented before any game needed it and has survived none. Imposing it on GROUND would manufacture the first use rather than discover it. So K14 states what is actually guaranteed, CommitWindow is marked provisional in the source, and it carries a delete-by date of 2026-12-31. Kernel spec->code link 16/18 -> 18/18, stated with the caveat the gate prints every run: that is about names, not assertions. Two self-tests broke and both broke correctly. rule-coverage's gate test hardcoded "unlinked rules exist today" and failed when the last one was linked; it now computes that and asserts the gate fails iff rules are unlinked. facts' text check rejected k_unlinked once it became legitimately empty; empty now renders as "(none)" and the check distinguishes absent from empty. M-D1-MUT: 8 of 14, unchanged — K14 and K18 are kernel rules, not acceptance rows. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-01 12:47:16 +02:00
check("kernel: gate never fails before the binding date",
before == 0, f"before={before}")
check("kernel: after the binding date the gate fails iff rules are unlinked",
after == (2 if _unlinked else 0),
f"after={after}, {len(_unlinked)} unlinked"
+ (f" ({' '.join(_unlinked)})" if _unlinked else " — all linked"))
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
print("rule-coverage self-test (positive control)")
ok = True
for name, passed, detail in results:
print(f" [{'ok ' if passed else 'FAIL'}] {name}"
+ (f"{detail}" if detail else ""))
ok &= passed
return 0 if ok else 1
def main():
2026-07-31 10:13:52 +02:00
# T01: inputs are repo-relative, so anchor to the repo rather than
# requiring the caller to `cd` first.
enter_root()
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
if "--self-test" in sys.argv:
return self_test()
rules = parse_rules(open("specs/GroundRules.md").read())
paths = sorted(glob.glob("scenarios/ground/*.yaml"))
# Positive control: refuse to report a percentage over nothing.
if not rules:
print("ERROR — no GR-rules parsed from specs/GroundRules.md; "
"refusing to report coverage", file=sys.stderr)
return 1
if not paths:
print("ERROR — no scenarios found in scenarios/ground/; "
"refusing to report coverage", file=sys.stderr)
return 1
covered = set()
for path in paths:
covered |= parse_covers(open(path).read())
known = set(rules)
hit = sorted(known & covered)
missing = [r for r in rules if r not in covered]
invented = sorted(covered - known)
# T09: the spec -> code -> scenario chain, made mechanical. A rule a
# scenario claims should also be named in the aggregate, or the claim
# rests on nothing but a tag.
code_ids = parse_code_ids(open(AGGREGATE).read())
unlinked = sorted((known & covered) - code_ids)
phantom = sorted(code_ids - known)
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
pct = 100 * len(hit) // len(rules)
linked = len((known & covered) & code_ids)
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
print(f"AM-1 rule coverage: {len(hit)}/{len(rules)} ({pct}%) "
f"over {len(paths)} scenarios")
print(f"AM-1b spec->code link: {linked}/{len(hit)} claimed rules also "
f"named in {AGGREGATE}")
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
print(" NOTE: counts tags; does not prove a scenario exercises what it names")
if unlinked:
print(" unlinked (claimed by a scenario, absent from the aggregate):")
print(" ", " ".join(unlinked))
if phantom:
print(" ERROR — rule id in code that the spec does not define:",
" ".join(phantom), file=sys.stderr)
return 1
# T08: provisional items are reported with an owner and an age.
prov = provisional_items(paths)
if prov:
import datetime
today = datetime.date.today()
print(f"\nprovisional U-item defaults: {len(prov)}")
unowned, stale = [], []
for path, owner, raised in prov:
age = "?"
if raised:
try:
age = (today - datetime.date.fromisoformat(raised)).days
except ValueError:
age = "?"
name = path.split("/")[-1]
print(f" {name:<34} owner={owner or 'NONE':<12} age={age}d")
if not owner:
unowned.append(name)
if isinstance(age, int) and age > PROVISIONAL_WARN_DAYS:
stale.append(f"{name} ({age}d)")
if unowned:
print(" WARN — provisional with no owner:", " ".join(unowned))
if stale:
print(f" WARN — provisional for over {PROVISIONAL_WARN_DAYS} days:",
" ".join(stale))
print(" NOTE: evidence files must list these; a ruling flips the "
"scenario, not the kernel")
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
if missing:
print(" uncovered:", " ".join(missing))
if invented:
print(" ERROR — claimed but not defined in the spec:", " ".join(invented),
file=sys.stderr)
return 1
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
kernel_rc = kernel_arm()
if kernel_rc:
return kernel_rc
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
return 0 if not missing else 2
CB-WP-0005 T01: spec->code link over every numbered spec and every crate AGGREGATE becomes a list of source roots and rule patterns become per-spec, so the link runs over every numbered spec x every crate rather than GroundRules.md x games/ground/src/lib.rs. The prediction held on the first run: AM-1b kernel spec->code link: 15/18 (83%) across 10 source files unlinked: K10 K14 K18 Kernel rules are link-only by design, and the output says so: they are kernel invariants with no aggregate, setup preset or command vocabulary, so scenarios/kernel/*.yaml with covers: [K11] would be a tag in a directory the runner cannot dispatch. Claiming scenario coverage for them is the inflation this gate exists to prevent. Per ADR-0005 §5 the kernel arm reports without feeding the exit code until 2026-08-31, then binds — the date in the tool, not in prose, with days remaining printed every run, because open-ended "gate it later" is how AM-4's targets went unratified for four workplans. The self-test asserts the gate returns 0 before that date and 2 after. The zero-rules positive control is replicated on the new denominator: a kernel regex that stops matching aborts rather than printing 0/0 as though it were 100%. The self-test passed while the tool was completely broken. A print( inside say() became say(), so every real `make coverage` died with RecursionError while --self-test reported all-ok — it only ever called kernel_arm(quiet=True) and never executed the reporting path. The control named the behaviour and did not assert it, which is precisely what this workplan is about. Fixed by exercising the loud path and asserting it prints, then verified by re-breaking say() and confirming both new checks go red. Seventh instance of the harness-does-nothing shape, in the tool written to find that shape. Also caught by its own gate: a self-test label that printed "0 K-ids" beside a passing ">5" assertion, because the detail string rebuilt the pattern with different escaping. A label that contradicts its own check is worse than no label. k_rules, k_linked and k_unlinked are registered facts under facts-check. A limit of that checker is recorded rather than patched: it is line-based, so a tagged value that prose-wraps fails. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 16:54:07 +02:00
def kernel_arm(today=None, quiet=False):
"""AM-1b over specs/GameKernel.md × every crate (CB-WP-0005 T01).
Link only see the module docstring for why K-rules cannot use the
scenario `covers:` mechanism. Returns a non-zero code only once
KERNEL_GATES_FROM has passed.
"""
today = today or datetime.date.today()
def say(*a, **kw):
if not quiet:
print(*a, **kw)
spec = os.path.join("specs", "GameKernel.md")
rules = parse_rules(open(spec).read(), r"\*\*(K\d+)\*\*")
# The inherited positive control. A pattern that stops matching must
# abort, not report 0/0 as though it were an answer.
if not rules:
print(f"\nERROR — no K-rules parsed from {spec}; refusing to report "
f"kernel coverage", file=sys.stderr)
return 1
paths = source_files()
if not paths:
print("\nERROR — no source files found under "
f"{'/, '.join(SOURCE_ROOTS)}/; refusing to report kernel coverage",
file=sys.stderr)
return 1
named = code_ids_over(paths, r"\bK\d+\b")
known = set(rules)
linked = sorted(known & named)
unlinked = [r for r in rules if r not in named]
phantom = sorted(named - known)
days = (KERNEL_GATES_FROM - today).days
binding = days <= 0
pct = 100 * len(linked) // len(rules)
say(f"\nAM-1b kernel spec->code link: {len(linked)}/{len(rules)} ({pct}%) "
f"K-rules named across {len(paths)} source files")
say(" NOTE: link only — K-rules are kernel invariants with no scenario "
"mechanism; and this counts names, not assertions (see "
"`make mutation-check`)")
if binding:
say(f" gate: BINDING since {KERNEL_GATES_FROM}")
else:
say(f" gate: reporting only for {days} more day(s), binds "
f"{KERNEL_GATES_FROM} (ADR-0005 §5)")
if unlinked:
say(" unlinked (declared in the spec, named nowhere in source):")
say(" ", " ".join(unlinked))
if phantom:
say(" ERROR — K-id in code that the spec does not define:",
" ".join(phantom), file=sys.stderr)
return 1
if unlinked and binding:
return 2
return 0
T01: audit every InnerLoop rule, and make the checkable ones executable 41 rules classified executable / checkable / decorative, each tagged with the failure class it catches. Counts: 11 executable, 22 checkable, 4 decorative (one of them dead policy). Audit: history/260731-inner-loop-rule-audit.md New tools/loop-lint.py makes 7 rules executable (tier declared, chaos roll recorded, tier-L review trail, unmeasured-in-evidence, whole-file loadability, reporting tools expose --self-test). It found three real violations on its first run, none previously visible: - specs/ArchitectureBlueprint.md was 543 lines against a ~400 limit the loop has stated since v0.2 and never measured. Split at its own section boundaries into Blueprint (1-8) + Runtime (9-15). - tools/dep-weight.py and tools/rule-coverage.py had positive-control logic and no --self-test, so nothing verified the control worked. Adding rule-coverage's self-test exposed a latent instance of the exact class this workplan is about: if the spec regex stopped matching, rules was empty, missing was empty, and the tool exited 0 reporting "0/0" -- a silent pass, in the tool that reports our headline AM-1 number. Both tools now assert they found something before reporting. Two demotions applied in the spec rather than left implicit: "structured over prose" is marked guidance (nothing can check it), and the 8k/10k token budget is struck through and marked DEAD POLICY pointing at T05. The audit's uncomfortable finding: rule 13 (re-derive inherited numbers) has no mechanical form, is deliberately left decorative, and caught the LARGEST error in CB-WP-0002. That is a counter-example to this workplan's own hypothesis. "A rule that cannot be executed is not a rule" is wrong as stated; the defensible version is that such a rule cannot be relied on to fire, so it must not be the only defence for a class that matters. Class coverage: harness-does-nothing has five executable rules; trusted-arithmetic has ZERO and produced the largest single error. make loop-lint and make self-tests wired into `make all` and CI. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-31 09:16:00 +02:00
if __name__ == "__main__":
sys.exit(main())