pest --parallel emits a single JSON line {"tool":"pest","tests":N,"passed":N,"skipped":N,...}
instead of human-readable text; the old regex-only parser returned 0/0/0sk/0 for every
parallel run. Added JSON-first branch with regex fallback; 3 new unit tests cover the
JSON path (passed+skipped, with failures, no skipped field).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Уровень «экономия 5%» = «0% без избыточности»: то же качество, что 0%,
вырезаны 6 пунктов дублирующей/бесполезной работы (re-read CLAUDE.md,
тесты-после-каждой-правки, gitleaks-full-history per-commit, Stop-верификатор,
авто-гейты brainstorming/plan-mode -> §12.2-floor). Уровень 0% не меняется.
cspell-words.txt: +коммитятся (валидная форма семейства коммит*).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Both tools check RLS compliance; the boundary "когда какой" was
undocumented (tracked as a RED conflict on the automation graph).
- .claude/skills/rls-check/SKILL.md: +section "Граница с агентом
rls-reviewer", +bullet in "Не использовать когда", +clause in
the frontmatter description.
- .claude/agents/rls-reviewer.md: +mirrored section "Граница со
скилом /rls-check", +bullet in "Out of scope", +clause in
the description.
- docs/automation-graph.html: conflict sk_rls<->ag_rls recolored
RED->GREEN (CONFLICT edge + both nd() node entries + EDGE_META).
- cspell-words.txt: +1 pre-existing word surfaced by the cspell
full-file scan of the now-staged SKILL.md.
Rule: one named table -> /rls-check; diff/branch/PR -> rls-reviewer.
The smoke test stays skill-only by design (running Pest in a review
subagent is slow and hits --parallel quirks 72/77).
Spec: docs/superpowers/specs/2026-05-16-rls-tooling-boundary-design.md
Plan: docs/superpowers/plans/2026-05-16-rls-tooling-boundary.md
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
A missing cmd-based tool on Windows exits 1 with an "is not recognized"
message, not POSIX exit 127. runCheck now also matches that message so a
missing composer/npm is classified SKIPPED (verdict RED-INCOMPLETE) per
spec §8, instead of a plain failure. Code-review follow-up for Task 7.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review fixups: Number.isFinite-guard в amountError/canSubmit
(очищенное number-поле не должно включать кнопку); watch(model) сбрасывает
amount/errorMsg при открытии (паттерн ReminderDialog, нет префилла/race);
комментарий про намеренный refetch в onTopupSuccess; flushPromises в spec.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
8-task plan for the rls-check skill <-> rls-reviewer agent boundary:
mirrored "Граница..." sections in both tool files, conflict recolor
RED->GREEN on the automation graph (4 spots), lint sweep, Playwright
visual smoke, one atomic commit, memory sync.
Spec: docs/superpowers/specs/2026-05-16-rls-tooling-boundary-design.md
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
4-task plan for iter6 of docs/automation-graph.html: «Паспорт узла»
legend section (since/changed/uses) for all 83 nodes + 2 footer toggle
buttons (usage heatmap, duplicate highlight). NODE_META (83 records) and
DUPLICATE_GROUPS (6 pairs D1-D5/D7) carry factual values derived from
76 session transcripts (window 09-16.05.2026) + git history; method and
raw outputs in Appendix A. cspell-words.txt += pcreator, pvalid.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review fixups: тест на :disabled PDF-кнопки по has_pdf
(spec-mandated поведение без покрытия); doc-комментарий billingFormatters
дополнен InvoicesTable в списке потребителей.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Дизайн-спек устранения конфликта 🔴 RED #1 карты автоматизации:
скил /rls-check и агент rls-reviewer оба проверяют RLS без чёткой
границы «когда какой».
Решение (Подход 1 — асимметрия как граница): оставить оба, прописать
регламент. Скил — одна названная таблица + живой дымовой тест;
агент — diff/ветка/PR, только 7 статических проверок. Дымовой тест
намеренно вне агента (Pest в ревью-субагенте медленный + задевает
квирки 72/77).
Затрагивает только проектно-локальные файлы инструментов + карту —
0 правок нормативки (Pravila/CLAUDE.md/PSR_v1/Tooling).
cspell-words.txt: +скиле +скилом (падежные формы термина «скил»).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review fixups: loadWallet() catch-блок сбрасывает wallet в
null (нет ложного рендера устаревших данных при неудачном повторе);
тест на кнопку «Повторить» (re-fetch + переход в success-состояние).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
api/billing.ts (getWallet) + BillingView тянет GET /api/billing/wallet
на mount (шапка + BalanceCard, loading/error-state). BalanceCard на
реальные props с nullable-тарифом. featureLabel для feature-слагов.
Sprint 2 Plan C, audit E3 (frontend pt1).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Spec §10 claimed run.mjs needs no unit harness, on the false premise that tools/*.mjs have no tests. In fact all 3 tools/*.mjs have a co-located .test.mjs (node:test). Amended §2/§3/§4/§10 + header note: run.mjs is split into exported pure functions (parsers, verdict, canonical-line, platform fork) + orchestrator, with a co-located run.test.mjs (node:test, ruflo-queen-hook.test.mjs pattern) — pure functions unit-tested, main subprocess-tested.
Aligns the spec with the economy-0% TDD mandate and the project tools/*.mjs convention before writing the implementation plan.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review fixups: runway_days клампится в 0 при отрицательном
балансе (overdrawn-тенант не должен показывать «−N дней»); (int)-каст в
wallet() для консистентности; усилены assertJsonPath на type-фильтре.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Brainstorming-phase design for custom skill #1 from claude-automation-recommender: a /regression skill packaging the project regression sweep (Pest --parallel, Vitest, Larastan, vue-tsc, lint/format, lychee, gitleaks) into one invocation — two tiers (quick/full), bundled .mjs orchestrator, canonical status line, GREEN/RED exit-code verdict.
Q1-Q4 design forks approved via brainstorming; spec self-review passed. cspell-words.txt: +6 project glossary transliterations introduced by the spec. Next: superpowers:writing-plans for the implementation plan.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review fixups: документирующий комментарий про безопасность
Eloquent save() для bcmath-строки (расхождение с LedgerService raw-update);
cross-tenant isolation тест на /api/billing/topup; balance_rub_after в
assertDatabaseHas.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
BillingTopupService кредитует tenants.balance_rub (bcmath) и пишет
append-only строку balance_transactions(type='topup'). BillingController
+ route POST /api/billing/topup под [auth:sanctum, tenant]. MVP-stub:
без платёжного шлюза (ЮKassa — post-Б-1).
Sprint 2 Plan C, audit E1 (backend).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
5-task план реализации audit-эпиков E1 (TopupDialog + POST
/api/billing/topup stub) и E3 (BillingView Overview на real API:
wallet/transactions/invoices). Backend: BillingController +
BillingTopupService + TariffPlan. Frontend: api/billing.ts + 4
компонента биллинга с mock на real API.
Sprint 2 Plan C. Источник: docs/superpowers/specs/2026-05-15-portal-audit-design.md.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
План реализации iter5 поверх spec dd20b4b: 2 задачи (реколлаж кластера
ruflo в automation-graph.html одним атомарным коммитом + синхронизация
memory). Полное литеральное содержание узлов/рёбер/деталей, верификация
через grep + visual smoke. +2 слова в cspell-words.txt (арг, греп).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review of Task 5: adds tests for the loadApiKey/loadWebhook
catch branches (apiKeyError/webhookError -> error v-alert) and changes
the Copy-button disabled check to the idiomatic falsy form.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Audit D2/D3/D4/D5: all four ApiTab buttons were handler-less and the
fields were hardcoded. Adds api/apiKeys.ts + api/webhooks.ts modules and
rewires ApiTab: loads the api-key prefix + webhook settings on mount;
Copy -> clipboard + snackbar; Regenerate -> confirm dialog -> POST
regenerate (full key shown once); Save Webhook -> PUT webhook-settings;
Test Webhook -> POST test with the result in a snackbar.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review of Task 4: adds a cross-tenant isolation test
(verifies the where(tenant_id) guard, matching ApiKeyControllerTest)
and a test()-endpoint failure-path test (HTTP 500 -> ok=false). Drops
the @return docblock from OutboundWebhookSubscriptionFactory for
consistency with ApiKeyFactory, eliminating a baseline entry at source.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Audit J5/D4/D5: the outbound_webhook_subscriptions table existed in
schema but had zero code. Adds the OutboundWebhookSubscription model +
factory and WebhookSettingsController with GET/PUT
/api/tenants/me/webhook-settings (one subscription per tenant; secret
generated + returned once on creation, bcrypt-hashed) and POST
/api/webhooks/test (unsigned connectivity check — HMAC-signed event
delivery is a separate post-MVP epic). Tenant-scoped via auth:sanctum +
tenant middleware.
phpstan-baseline.neon: additive-only entries for new test file
(Pest\PendingCalls\TestCall false-positives — documented project pattern)
and OutboundWebhookSubscriptionFactory method.childReturnType (same
pattern as ProjectFactory/TenantFactory/UserFactory already in baseline).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review of Task 3: index() filtered by is_active only —
an expired-but-active key would be listed as valid. Adds an
expires_at > now() filter plus a test. Cannot occur today (regenerate
is the only write path, always +1 year) but is the correct semantic
contract for an «active key» listing.
phpstan-baseline.neon: count bumps only for ApiKeyControllerTest.php
($tenant 5→7, $user 3→5, getJson 3→4).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Audit J5/D3: the api_keys table existed in schema but had zero code.
Adds the ApiKey model + factory, and ApiKeyController with GET
/api/api-keys (list active keys, key_hash hidden) and POST
/api/api-keys/regenerate (deactivate prior + create new, full key
returned once, bcrypt-hashed in DB). Tenant-scoped via auth:sanctum +
tenant middleware (RLS on api_keys). phpstan-baseline.neon updated for
Pest PendingCalls false-positives in the new test file; also removes
8 pre-existing stale ignore.unmatched entries (properties now resolved
by existing @mixin IdeHelper* docblocks — confirmed pre-existing via
git stash test before Task 3 changes).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review of Task 2: documents why ProfileTab needs no
watch-resync of auth.user (router beforeEach awaits fetchMe before
requiresAuth navigation); tightens the save-error test to assert the
exact fallback message instead of mere truthiness.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Audit D1: ProfileTab fields were hardcoded refs and the Save button had
no handler. Rewired to the auth store + a new api/auth updateProfile()
calling PATCH /api/auth/me. Single «Полное имя» field split into Имя +
Фамилия (matches users.first_name/last_name); decorative «Роль» field
removed (no such column). AuthUser type gains phone + timezone.
SettingsView.spec.ts updated: «Полное имя» assertion changed to check
for «Имя» and «Фамилия» (collateral fix for the intentional field split).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review of Task 1: first_name had a 422 test but last_name
(identical required rule) did not. Adds the symmetric test.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
SyncSupplierProjectsJob:77 has a time-budget guard that breaks the
sync loop after 20:55 Europe/Moscow. Five of the eight tests in
SyncSupplierProjectsJobTest omitted Carbon::setTestNow(), so they
inherited real wall-clock time and silently failed (job no-ops)
every evening after 20:55 MSK -- a latent test bug since dedaae5
(Plan 3), mis-attributed to a Redis race (quirk 72) in earlier audits.
Pins beforeEach to a fixed pre-cutoff clock; the job code is correct
and unchanged. Verified: 8/8 in isolation, full suite back to green.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Audit J6: ProfileTab needs a full-profile update endpoint. Adds
AuthController::updateProfile (first_name/last_name/phone/timezone),
routed in the existing /api/auth auth:sanctum group; mirrors the
sibling updateNotificationPreferences. userResource() now also returns
phone + timezone so the GET /me round-trip carries them.
phpstan-baseline.neon updated for Pest PendingCalls false positives
in the new test file (same pattern as all other Feature test files).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan B of the Sprint 2 split — the Settings subsystem, 5 atomic TDD
tasks: PATCH /api/auth/me profile endpoint (J6); ProfileTab rewired to
real API (D1); ApiKey model + api-keys endpoints (J5/D3); outbound
webhook settings endpoints (J5/D4/D5); ApiTab full wiring (D2-D5).
Schema delta = 0 — api_keys + outbound_webhook_subscriptions tables
already exist in schema.sql.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review of the legal stub pages: the always-present
informational v-alert defaulted to role=alert (assertive live-region) —
changed to role=note for a static advisory (WCAG 2.1 AA). Trimmed
cosmetic whitespace inside the back-link element.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Audit A7: the «Оферта» / «Политика» links in the AuthLayout footer were
raw <a href> pointing at unrouted paths -> 404 via the SPA catch-all.
Adds a single DRY LegalDocView served by /legal/:doc(offer|privacy),
rendering an honest «document being finalized» stub (real legal text
needs юр. редактура — реестр K3 / blocker Б-1). Footer links upgraded
to <RouterLink> for SPA navigation. Also refreshes two stale auth-layout
doc-comments left by the /recovery removal (review M1).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Audit A2/A3: RecoveryCodesView (route /recovery) had a TODO no-op
continue handler and 8 hardcoded mock codes. Recon found the page is
orphaned — nothing in the UI navigates to /recovery. The real 2FA
recovery-codes flow lives entirely in Settings -> Безопасность
(TwoFactorCard setup wizard + RecoveryCodesCard regeneration), both
already wired to the real API. Per user decision (2026-05-15) the
orphan is deleted rather than polished.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Дмитрий