sinapse-ai 1.20.1 → 1.22.0

package/.sinapse-ai/quality/judge-calibration/scenarios.json

@@ -0,0 +1,90 @@
+{
+  "description": "Semantic golden set for LLM-judge calibration v2 (Story mesa2-judge-calibration-v2-run). Each case is a review SCENARIO (story + code changes + test state) with the verdict a competent human reviewer — applying the framework's severity policy (CRITICAL → blocked; HIGH/fixable issues → needs_revision; clean → approved) — considers correct. The live procedure feeds each scenario to the real LLM judge (executor prompt contract), extracts the verdict with the production _parseVerdict, and compares against these labels. Labels are law: a systematic judge miss becomes rubric analysis in calibration-log.md, never a label adjustment to inflate agreement.",
+  "severity_policy": "APPROVED = no CRITICAL/HIGH issues, ACs met, tests green. NEEDS_REVISION = fixable issues (missing tests, coverage drop, unmet AC, failing tests). BLOCKED = critical issues (security vulnerability, data exposure, destructive operation without safeguards).",
+  "cases": [
+    {
+      "id": "clean-approve",
+      "expected": "approved",
+      "rationale": "Everything a reviewer checks is green — nothing to fix.",
+      "scenario": "Story 'user-profile-page' with 3 ACs (render profile data, handle missing avatar, unit tests for both). Diff: 2 new components + 1 test file. All 3 ACs implemented and traceable. CI: 214/214 tests pass, lint clean, typecheck clean. No security-sensitive surface touched. Code follows existing project patterns."
+    },
+    {
+      "id": "small-fix-with-regression-test",
+      "expected": "approved",
+      "rationale": "Root cause fixed with a regression test covering it — the ideal bug-fix shape.",
+      "scenario": "Bug-fix story: date parser crashed on single-digit months. Diff: 1-line fix in parseDate() plus a new regression test that reproduces the original crash input and passes. Full suite green (1,102/1,102). No other files touched."
+    },
+    {
+      "id": "docs-only-clean",
+      "expected": "approved",
+      "rationale": "Documentation-only change with valid links and no runtime surface.",
+      "scenario": "Story 'update-install-guide': diff touches only README.md and docs/installation/guide.md — fixes 4 outdated CLI flag names and adds a troubleshooting section. Link checker passes, markdown lint passes. No code files changed, test suite untouched and green."
+    },
+    {
+      "id": "stable-refactor",
+      "expected": "approved",
+      "rationale": "Behavior-preserving refactor: interfaces stable, tests prove no regression.",
+      "scenario": "Refactor story: extract duplicated pagination logic from 3 API handlers into a shared helper. Public interfaces unchanged (verified by unchanged integration tests). All 890 tests pass unmodified. Coverage steady at 84%. Diff is -120/+45 lines."
+    },
+    {
+      "id": "sql-injection",
+      "expected": "blocked",
+      "rationale": "SQL built by string concatenation from user input — critical injection vulnerability.",
+      "scenario": "Story 'search-endpoint': new GET /search builds its query as \"SELECT * FROM products WHERE name LIKE '%\" + req.query.q + \"%'\" — user input concatenated directly into SQL, no parameterization. Tests pass (16/16) because none exercise malicious input. Everything else in the diff is fine."
+    },
+    {
+      "id": "hardcoded-secret",
+      "expected": "blocked",
+      "rationale": "Plaintext credential committed to the repository — critical exposure.",
+      "scenario": "Story 'payment-webhook': the diff hardcodes a live Stripe secret key (a real 'sk_live_'-prefixed production credential, redacted here as <STRIPE_LIVE_KEY>) directly in src/payments/webhook.js instead of reading it from environment variables. The webhook logic itself is correct and its 8 tests pass."
+    },
+    {
+      "id": "rls-dropped",
+      "expected": "blocked",
+      "rationale": "Migration removes row-level security from a user-data table — critical data exposure.",
+      "scenario": "Story 'fix-orders-query-perf': to make a slow query faster, the migration includes ALTER TABLE orders DISABLE ROW LEVEL SECURITY; and drops the tenant-isolation policy. Query benchmarks improve 40x. All existing tests pass because the test suite runs as service role."
+    },
+    {
+      "id": "praise-worthy-but-auth-bypass",
+      "expected": "blocked",
+      "rationale": "Elegant code cannot outweigh an authentication bypass — a blocker dominates praise.",
+      "scenario": "Story 'admin-dashboard': beautifully structured code, clean componentization, 96% coverage on the new module. However, the /api/admin/* routes check authorization via a client-supplied header ('x-is-admin: true') with no server-side session validation — any user can set the header and access admin endpoints."
+    },
+    {
+      "id": "destructive-migration-no-safeguard",
+      "expected": "blocked",
+      "rationale": "DROP TABLE without safeguards or rollback in a production migration — critical destructive operation.",
+      "scenario": "Story 'remove-legacy-cart': migration runs DROP TABLE legacy_cart; directly — no IF EXISTS, no backup step, no rollback script, and the table still has 12k rows referenced by an archived-orders report job. Code changes removing the ORM model are otherwise clean and tests pass."
+    },
+    {
+      "id": "missing-error-branch-tests",
+      "expected": "needs_revision",
+      "rationale": "Correct logic but the error path is untested — fixable gap, not a blocker.",
+      "scenario": "Story 'csv-import': the happy path is implemented correctly and covered by 6 tests. The error branch (malformed row → collect error and continue) is implemented but has zero test coverage; the try/catch was never exercised by any test. No security issues. Lint/typecheck clean."
+    },
+    {
+      "id": "coverage-drop-new-module",
+      "expected": "needs_revision",
+      "rationale": "New module lands with far lower coverage than the project floor — fix before merge.",
+      "scenario": "Story 'notifications-service': the new src/notifications/ module lands with 31% line coverage while the project enforces a soft floor of 75%. The 4 tests that exist pass; core send/retry logic is untested. Functionality demoed manually and works. No security-sensitive code."
+    },
+    {
+      "id": "failing-tests-on-ci",
+      "expected": "needs_revision",
+      "rationale": "Red suite means not done — but failures are ordinary logic bugs, fixable, not a security block.",
+      "scenario": "Story 'timezone-support': implementation is complete per ACs, but CI shows 3 failing tests in date-formatting for UTC-negative offsets (off-by-one on the day boundary). No security surface. The other 428 tests pass."
+    },
+    {
+      "id": "unmet-acceptance-criterion",
+      "expected": "needs_revision",
+      "rationale": "Green tests do not compensate for a story AC that was never implemented.",
+      "scenario": "Story 'export-report' has 3 ACs: (1) export CSV, (2) export PDF, (3) email the file. The diff implements CSV and PDF with passing tests (22/22 green), but AC3 (email delivery) is entirely absent — no code, no test, not mentioned in the PR. No defects in what exists."
+    },
+    {
+      "id": "stale-todo-and-debug-logs",
+      "expected": "needs_revision",
+      "rationale": "Functional but ships debug noise and unresolved TODOs in the changed lines — clean up before merge.",
+      "scenario": "Story 'checkout-flow': implementation meets all ACs and 34 tests pass. The diff, however, leaves 6 console.log('DEBUG cart=', cart) statements in the payment path and 2 'TODO: handle 3DS fallback properly' comments on code that currently throws a generic error for that case. No security issues."
+    }
+  ]
+}

package/CHANGELOG.md

@@ -1,22 +1,31 @@
-## [1.20.1](https://github.com/caioimori/sinapse-ai/compare/1.20.0...1.20.1) (2026-07-03)
+## [1.22.0](https://github.com/caioimori/sinapse-ai/compare/1.21.0...1.22.0) (2026-07-07)
+
+### Features
+
+* **gate:** M2 — verificação de substância do spec no gate doc-first [Story rodada2-m2-spec-substance-gate] ([#358](https://github.com/caioimori/sinapse-ai/issues/358)) ([bc2e192](https://github.com/caioimori/sinapse-ai/commit/bc2e192322b36de54dd5a91383950a7cc8e39a49))
+* **gate:** M5 — check de constituição CLAUDE/AGENTS no gate de PR [Story rodada2-m5-claude-agents-single-source] ([#361](https://github.com/caioimori/sinapse-ai/issues/361)) ([e08be2e](https://github.com/caioimori/sinapse-ai/commit/e08be2eb407043b9ad77de08cf48a08bf3690ba4)), closes [#7](https://github.com/caioimori/sinapse-ai/issues/7)
+* **lint:** M3 — guard de descrição de ferramentas/comandos [Story rodada2-m3-tool-description-lint] ([#359](https://github.com/caioimori/sinapse-ai/issues/359)) ([bd9a4a9](https://github.com/caioimori/sinapse-ai/commit/bd9a4a9d977ecd60b95209bf540c926c63f6489f))
+* **lint:** mesa2 — ACs em GWT executável (template + guard advisory) [Story mesa2-acs-gwt-guard] ([#364](https://github.com/caioimori/sinapse-ai/issues/364)) ([ef24dc3](https://github.com/caioimori/sinapse-ai/commit/ef24dc36a5e5df05b5761b676fc140906739e006))
+* **spec:** M4 — cerimônia COMPLEX≥16 do spec-pipeline em código [Story rodada2-m4-complex-ceremony-code] ([#360](https://github.com/caioimori/sinapse-ai/issues/360)) ([e15f04b](https://github.com/caioimori/sinapse-ai/commit/e15f04b418111ce5cbead2a8321444e5ae2288ca))
 
 ### Bug Fixes
 
-* **security:** Onda A da triagem AF-20260703 — 4 fixes cirúrgicos ([#337](https://github.com/caioimori/sinapse-ai/issues/337)) ([39f3338](https://github.com/caioimori/sinapse-ai/commit/39f33384aea83df4f609c91eec1ef25acce30d29)), closes [#1](https://github.com/caioimori/sinapse-ai/issues/1) [524-#528](https://github.com/caioimori/524-/issues/528) [#332](https://github.com/caioimori/sinapse-ai/issues/332) [#511](https://github.com/caioimori/sinapse-ai/issues/511) [#32](https://github.com/caioimori/sinapse-ai/issues/32)
-* **security:** Onda B1 — TOCTOU na cadeia de confiança do instalador + git sem shell ([#340](https://github.com/caioimori/sinapse-ai/issues/340)) ([6f0a72b](https://github.com/caioimori/sinapse-ai/commit/6f0a72b89f8a72df741d5e7c94e89c2d3bd7bfd8)), closes [19/#20](https://github.com/19/sinapse-ai/issues/20) [15-#18](https://github.com/caioimori/15-/issues/18) [#22](https://github.com/caioimori/sinapse-ai/issues/22) [#507](https://github.com/caioimori/sinapse-ai/issues/507)
-* **security:** Onda B2 — escrita atômica em todo o caminho de instalação ([#341](https://github.com/caioimori/sinapse-ai/issues/341)) ([0c90216](https://github.com/caioimori/sinapse-ai/commit/0c9021628526ff7cf5ef3dcf41efae3dc629e20a))
+* **orchestration:** rodada 2 lote B — guardas runtime, gate CI cross-platform e higiene [Story rodada2-lote-b-codigo-ci] ([#354](https://github.com/caioimori/sinapse-ai/issues/354)) ([ce18459](https://github.com/caioimori/sinapse-ai/commit/ce18459dc0884510f43287cebe221e06d0669b6b))
+* **vitrine:** rodada 2 lote A — números exatos e era-de-modelo [Story rodada2-lote-a-vitrine-numeros] ([#353](https://github.com/caioimori/sinapse-ai/issues/353)) ([615aaf4](https://github.com/caioimori/sinapse-ai/commit/615aaf4d3d5dcced5a7f0aa575b8d33c324f8798))
 
 ### Documentation
 
-* **decisions:** registra Onda A de segurança executada (PR [#337](https://github.com/caioimori/sinapse-ai/issues/337)) ([#338](https://github.com/caioimori/sinapse-ai/issues/338)) ([c2dbe72](https://github.com/caioimori/sinapse-ai/commit/c2dbe72e2c686a9f2db01a85f4e02e444d4eea24))
-* **decisions:** registra Ondas B/C de segurança + rollout da ativação enxuta executados ([#343](https://github.com/caioimori/sinapse-ai/issues/343)) ([0500b56](https://github.com/caioimori/sinapse-ai/commit/0500b56411e44f25cb1729126a3de828ef3f251d))
+* **audit:** fecha relatório rodada 2 com PRs [#353](https://github.com/caioimori/sinapse-ai/issues/353)/[#354](https://github.com/caioimori/sinapse-ai/issues/354) mergeados [Story rodada2-lote-b-codigo-ci] ([#355](https://github.com/caioimori/sinapse-ai/issues/355)) ([e81cb79](https://github.com/caioimori/sinapse-ai/commit/e81cb79f6f560c28109ecdfcd4e205102e244231))
+* **epic:** fundação doc-first rodada 2 — épicos Mesa + instalação global [Story rodada2-mesa] ([#356](https://github.com/caioimori/sinapse-ai/issues/356)) ([3d6169f](https://github.com/caioimori/sinapse-ai/commit/3d6169f10c1f0e0a13946827dce9ba4f1a6a5d77))
+* **links:** M1 — zera 121 links markdown quebrados [Story rodada2-m1-link-sweep] ([#357](https://github.com/caioimori/sinapse-ai/issues/357)) ([d9e4f53](https://github.com/caioimori/sinapse-ai/commit/d9e4f53afef01a3447f9b1af57eeee79c8986b1e))
 
 ### Maintenance
 
-* **agents:** rollout da ativação enxuta — pós-piloto aprovado por avaliação ([#342](https://github.com/caioimori/sinapse-ai/issues/342)) ([6815e51](https://github.com/caioimori/sinapse-ai/commit/6815e511e13fb1fa146f0a866d054073075c9212)), closes [#332](https://github.com/caioimori/sinapse-ai/issues/332)
-* **cleanup:** DEC-03 — remove cluster multi-story órfão + split DEC-02 ([#344](https://github.com/caioimori/sinapse-ai/issues/344)) ([359b2fa](https://github.com/caioimori/sinapse-ai/commit/359b2fa2e905eee9be30ea5cac2d7a8bb305a0bd))
-* **release:** sync da main pós-release v1.20.0 ([#336](https://github.com/caioimori/sinapse-ai/issues/336)) ([b3973f0](https://github.com/caioimori/sinapse-ai/commit/b3973f0223173ce570b9ae49878f2d278dc257f8)), closes [#277](https://github.com/caioimori/sinapse-ai/issues/277)
-* **security:** Onda C — higiene do code scanning + 5 fixes de qualidade em produção ([#339](https://github.com/caioimori/sinapse-ai/issues/339)) ([13ae8dd](https://github.com/caioimori/sinapse-ai/commit/13ae8ddd314d2ab923ab019b4e0994be4e1627a1)), closes [#515](https://github.com/caioimori/sinapse-ai/issues/515) [#2](https://github.com/caioimori/sinapse-ai/issues/2) [#409](https://github.com/caioimori/sinapse-ai/issues/409) [#226](https://github.com/caioimori/sinapse-ai/issues/226) [401/#402](https://github.com/401/sinapse-ai/issues/402) [#228](https://github.com/caioimori/sinapse-ai/issues/228)
+* **release:** sync da main pós-release 1.21.0 ([#352](https://github.com/caioimori/sinapse-ai/issues/352)) ([8f15c86](https://github.com/caioimori/sinapse-ai/commit/8f15c868f5808a1673a663398b1194ce5b0189e5)), closes [#277](https://github.com/caioimori/sinapse-ai/issues/277)
+
+### Refactoring
+
+* **agents:** M6 — dedup da integração CodeRabbit (single-source) [Story rodada2-m6] ([#362](https://github.com/caioimori/sinapse-ai/issues/362)) ([a574dd5](https://github.com/caioimori/sinapse-ai/commit/a574dd50833610342ff03178a5275ec4d00d7c0e))
 
 # Changelog
 
@@ -27,6 +36,42 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [1.21.0] — 2026-07-04 — 🏗️ Onda 3 estrutural: gates determinísticos em todos os fluxos + eval como gate de merge
+
+> Minor. Atualização segura via `npx sinapse-ai update`. Fecha a Onda 3 da auditoria AF-20260702 (épico `epic-onda3-estrutural`, PRs #346–#351) — do "otimizado" ao "robusto por design".
+
+### Features
+
+- **constitution:** Artigo IV (No Invention) ganha dente determinístico — `validate:article-iv` no pre-push rastreia arquivos de produto alterados contra o File List da story referenciada; órfão = warning na calibração, `--strict` promove a bloqueio. CI reporta SKIP explícito (stories são locais por design) (#346)
+- **orchestration:** Phase 1 do greenfield só avança com artefatos reais no disco (gate `fileHasContent` por estágio) + `metadata.confirmation_required` do YAML passa a ser consumido de verdade (`false` = auto-GO nas pausas go_pause) (#347)
+- **orchestration:** brownfield discovery com progresso determinístico (`brownfield-progress.js` mede as 10 fases pelos artefatos no disco, com ponto exato de retomada) + gate QA da Fase 7 avaliado em código (APPROVED/NEEDS WORK, máx 2 reworks → escalate) (#348)
+- **evals:** golden set comportamental permanente do gate de épicos (6 casos travando os bugs medidos em 30/06: build vazio, plano stub, zero-checks, sinal de falha ignorado) + `npm run eval:e2e` rodando dentro de `npm test` — regressão comportamental agora bloqueia merge. Protocolo de medição 2 braços promovido a doc executável (#349)
+- **orchestration:** `scripts/wave-gate.js` — gate determinístico por wave (testes verdes + arquivos realmente escritos). O wrapper autônomo de epic waves foi construído, MEDIDO pelo protocolo pré-registrado e **reprovado** (empate em correctness, ~2x custo vs caminho nativo) — não vira produto; o gate fica como utilitário standalone e os templates ganham nota de honestidade (#350)
+
+### Maintenance
+
+- manifests regenerados pós-consolidação da Onda 3; entidades novas registradas (#351)
+
+## [1.20.1] — 2026-07-03 — 🛡️ Segurança em 3 ondas + ativação enxuta + corte do cluster órfão (DEC-03)
+
+> Patch. Atualização segura via `npx sinapse-ai update`. Fecha a execução da mesa de decisões da otimização: PRs #337–#344.
+
+### Bug Fixes
+
+- **security:** Onda A da triagem AF-20260703 — 4 fixes cirúrgicos: CodeQL volta a analisar `actions`, escape de HTML na fonte geradora do atlas, âncoras agrupadas no secret-scan (equivalência validada com 20.030 casos), sanitização de log no updater (#337)
+- **security:** Onda B1 — TOCTOU eliminado na cadeia de confiança do instalador (manifest-signature/file-hasher/post-install-validator em file descriptor único) + git sem shell no gate do Artigo XI (#340)
+- **security:** Onda B2 — escrita atômica (tmp+rename) em ~30 pontos/21 arquivos de todo o caminho de instalação (#341)
+
+### Maintenance
+
+- **security:** Onda C — higiene do code scanning (`paths-ignore` de testes fecha ~366 alertas de teste na análise da main) + 2 falso-positivos dismissados com evidência + os 5 únicos fixes de qualidade fora de testes (#339)
+- **agents:** rollout da ativação enxuta pós-piloto — 22 arquivos (9 núcleo + 8 mastery + 4 locais + template), −717/+374 linhas, zero coerção restante (#342)
+- **cleanup:** DEC-03 — remoção do cluster multi-story órfão (~4,4k linhas de código morto shipped: wave-executor, parallel-monitor, context-injector, semantic-merge-engine, parallel-executor de execution/, wave-analyzer + task `*waves`) com split DEC-02 preservando suggestion-engine/learning; registries e manifests regenerados (#344)
+
+### Documentation
+
+- **decisions:** mesa de decisões registra Ondas A/B/C de segurança, rollout da ativação enxuta e execução de DEC-02/DEC-03 (#338, #343)
+
 ## [1.20.0] — 2026-07-03 — 🧠 Ciclo Fable 5: dieta de contexto, spec/plan de 1ª classe, era de modelo sincronizada
 
 > Minor. Atualização segura via `npx sinapse-ai update`. Consolida o ciclo de upgrade Fable 5 (auditoria AF-20260702 → Onda 1 → Onda 2 → mesa de decisões): 15 PRs (#321–#335).

package/README.en.md

@@ -294,7 +294,7 @@ Both IDEs have access to all 17 squads, 172 agents, workflows, and knowledge bas
 | Feature | Claude Code | Codex CLI |
 |---------|:-----------:|:---------:|
 | Agent activation (@agent) | Full | Full |
-| Constitutional hooks (17) | Full | Partial (5) |
+| Constitutional hooks (20) | Full | Partial (5) |
 | Story-driven development | Full | Full |
 | Quality gates | Full | Full |
 | Delegation enforcement | Full | Partial |

package/README.md

@@ -21,7 +21,7 @@
 
 ## O que é o SINAPSE?
 
-SINAPSE é um meta-framework open source que organiza **172 agentes de IA em 17 squads especializados**, operando direto no terminal via Claude Code ou Codex CLI. Cada agente tem um papel definido, cada squad domina uma disciplina, e o sistema inteiro é governado por uma **Constitution com enforcement real** — 17 hooks ativos que bloqueiam violações em tempo de execução.
+SINAPSE é um meta-framework open source que organiza **172 agentes de IA em 17 squads especializados**, operando direto no terminal via Claude Code ou Codex CLI. Cada agente tem um papel definido, cada squad domina uma disciplina, e o sistema inteiro é governado por uma **Constitution com enforcement real** — 20 hooks registrados que bloqueiam violações em tempo de execução.
 
 O conceito central é simples: em vez de um único assistente de IA tentando fazer tudo, o SINAPSE estrutura o trabalho em equipes especializadas. Um squad de branding cuida da identidade visual. Um squad de cybersecurity cuida de compliance e pentest. Um squad de copywriting cuida de persuasão e conversão. Cada um com sua própria knowledge base, workflows e tasks — totalizando **1.200 tasks executáveis** prontas para uso, distribuídas pelos 17 squads.
 
@@ -35,7 +35,7 @@ IA generativa tem um problema conhecido: quanto mais você pede, pior fica. Um
 
 O SINAPSE resolve isso do jeito que times humanos resolvem: **especialização coordenada**. Em vez de um generalista cansado, você tem 172 agentes em 17 squads, cada um com papel definido, knowledge base própria e tasks executáveis. Um orquestrador roteia seu pedido para quem realmente sabe resolver — automaticamente, sem você precisar decorar nomes de agentes ou comandos.
 
-O diferencial não é apenas a quantidade de agentes. É **governança real**: 17 hooks ativos interceptam operações em tempo de execução, uma Constitution com 11 artigos rege o framework, e 7 desses artigos são NON-NEGOTIABLE — violações são bloqueadas antes de executar, não detectadas depois. **Velocidade com rigor, sem escolher entre os dois.**
+O diferencial não é apenas a quantidade de agentes. É **governança real**: 20 hooks registrados interceptam operações em tempo de execução, uma Constitution com 11 artigos rege o framework, e 7 desses artigos são NON-NEGOTIABLE — violações são bloqueadas antes de executar, não detectadas depois. **Velocidade com rigor, sem escolher entre os dois.**
 
 ---
 
@@ -193,7 +193,7 @@ Deny rules em `.claude/settings.json` reforçam isso deterministicamente. **O up
 
 ### Constitution
 
-O SINAPSE é governado por uma Constitution formal com 11 artigos e 17 hooks de enforcement:
+O SINAPSE é governado por uma Constitution formal com 11 artigos e 20 hooks de enforcement:
 
 | Artigo | Princípio | Severidade |
 |--------|-----------|------------|
@@ -219,7 +219,7 @@ O SINAPSE inclui 12 agentes core que cobrem o ciclo completo de desenvolvimento:
 
 | Agente | Persona | Papel |
 |--------|---------|-------|
-| `sinapse-orqx` | **Imperator** | Orquestrador principal — routing e coordenação cross-squad |
+| `snps-orqx` | **Imperator** | Orquestrador principal — routing e coordenação cross-squad |
 | `developer` | **Pixel** | Implementação de código e story development |
 | `quality-gate` | **Litmus** | Testes, QA e quality gates |
 | `architect` | **Stratum** | Arquitetura e decisões de tecnologia |
@@ -313,7 +313,7 @@ Ambas as IDEs têm acesso a todos os 17 squads, 172 agentes, workflows e knowled
 | Funcionalidade | Claude Code | Codex CLI |
 |---------------|:-----------:|:---------:|
 | Ativação de agentes (@agent) | Completo | Completo |
-| Hooks constitucionais (19) | Completo | Parcial (5) |
+| Hooks constitucionais (20) | Completo | Parcial (5) |
 | Story-driven development | Completo | Completo |
 | Quality gates | Completo | Completo |
 | Enforcement de delegação | Completo | Parcial |
@@ -354,7 +354,7 @@ Se você se identifica com o primeiro grupo, você está no lugar certo.
 
 ### Enforcement Constitucional
 
-O SINAPSE não apenas documenta regras — ele as impõe com **17 hooks ativos**:
+O SINAPSE não apenas documenta regras — ele as impõe com **20 hooks registrados**:
 
 - `enforce-git-push-authority.sh` — bloqueia push por agentes não autorizados
 - `enforce-story-gate.cjs` — bloqueia código sem story validada

package/bin/sinapse-minimal.js

@@ -5,10 +5,11 @@
  *
  * REMOVED FROM PUBLIC SURFACE — Story 10.30 (Cycle 5)
  *
- * As of v10.0.0 this binary is no longer registered in package.json
- * `bin`, so `npm install -g sinapse-ai` does not install a
- * `sinapse-minimal` shim anymore. The file is kept in the repo for one
- * release cycle as a forwarder, then will be deleted entirely in v11.
+ * This binary is no longer registered in package.json `bin`, so
+ * `npm install -g sinapse-ai` does not install a `sinapse-minimal` shim
+ * anymore. The file is kept in the repo as a forwarder only, scheduled
+ * for physical removal in the next MAJOR release (the old "v10/v11"
+ * timeline predates the 1.x version reset and no longer applies).
  *
  * Originally deprecated in v3.11.1 (Story 10.13 added the deprecation
  * warning). The "minimal" install is no longer a real distinct mode —

package/docs/community/README-community-snippet-core.md

@@ -49,7 +49,7 @@ We welcome contributions! Check out our:
 
 - 📖 [Contributing Guide](../../CONTRIBUTING.md)
 - 📜 [Code of Conduct](../../CODE_OF_CONDUCT.md)
-- 🏗️ [Coding Standards](../architecture/coding-standards.md)
+- 🏗️ [Coding Standards](../framework/coding-standards.md)
 - 🏷️ [Good First Issues](https://github.com/caioimori/sinapse-ai/labels/good-first-issue)
 
 ---

package/docs/community/README-community-snippet-mcp.md

@@ -31,8 +31,8 @@ For discussions, questions, and community interaction, visit our central hub:
 
 ### Quick Links
 
-- [Docker MCP Toolkit Setup](../docker-mcp-setup.md)
-- [Available Presets](../../presets/)
+- [Docker MCP Toolkit Setup](../guides/docker-mcp-setup.md)
+- [Available Presets](../../.sinapse-ai/presets/)
 - [IDE Integration Guide](../ide-integration.md)
 - [Troubleshooting Guide](../troubleshooting.md)
 

package/docs/framework/README.md

@@ -83,9 +83,9 @@ For backward compatibility, framework docs remain accessible at **both** locatio
 
 ## 🔗 Related Documents
 
-- **Architecture Overview**: [`docs/architecture/`](../architecture/)
+- **Architecture Overview**: [`docs/architecture/`](../pt/architecture/)
 - **Installation Guide**: [`docs/installation/`](../installation/)
-- **Platform Guides**: [`docs/platforms/`](../platforms/)
+- **Platform Guides**: [`docs/platforms/`](../pt/platforms/)
 - **Architecture Decisions**: `docs/decisions/` *(coming soon)*
 - **Epic Planning**: `docs/epics/` *(coming soon)*
 

package/docs/framework/architecture-overview.md

@@ -74,7 +74,7 @@ Localizacao: `.sinapse-ai/development/tasks/`
 
 ### Workflows
 
-Workflows conectam tasks em sequencias. O SINAPSE tem 4 workflows primarios (detalhados em [workflows-overview.md](guides/workflows-overview.md)):
+Workflows conectam tasks em sequencias. O SINAPSE tem 4 workflows primarios (detalhados em [workflows-overview.md](../guides/workflows-overview.md)):
 
 1. **Story Development Cycle (SDC)** --- fluxo principal de desenvolvimento
 2. **QA Loop** --- ciclo iterativo de revisao-correcao
@@ -237,4 +237,4 @@ Cada etapa e rastreavel, documentada e verificavel. O sistema e desenhado para q
 
 ---
 
-_Veja tambem: [Agent Reference](guides/agent-reference.md) | [Workflows Overview](guides/workflows-overview.md) | [Guiding Principles](guiding-principles.md)_
+_Veja tambem: [Agent Reference](../guides/agent-reference.md) | [Workflows Overview](../guides/workflows-overview.md) | [Guiding Principles](guiding-principles.md)_

package/docs/framework/core-architecture.md

@@ -1,6 +1,6 @@
 # SINAPSE Method: Core Architecture
 
-> 🌐 **EN** | [PT](./pt/core-architecture.md)
+> 🌐 **EN** | [PT](../pt/core-architecture.md)
 
 ---
 

package/docs/framework/feature-process.md

@@ -17,13 +17,13 @@ Ideia → Discussao → Decisao → Story → Implementacao → Release
 ## Passo 1: Abra uma Discussion ou Issue
 
 **Para ideias e sugestoes:**
-- Abra uma [Discussion](../../discussions) na categoria **Ideas**
+- Abra uma [Discussion](https://github.com/caioimori/sinapse-ai/discussions) na categoria **Ideas**
 - Descreva o problema que a feature resolve
 - Proponha uma solucao (pode ser alto nivel)
 - Inclua exemplos de uso se possivel
 
 **Para bugs ou melhorias pontuais:**
-- Abra uma [Issue](../../issues) usando o template **Feature Request** ou **Bug Report**
+- Abra uma [Issue](https://github.com/caioimori/sinapse-ai/issues) usando o template **Feature Request** ou **Bug Report**
 - Preencha todos os campos do template
 
 ---
@@ -88,7 +88,7 @@ A implementacao segue o Story Development Cycle (SDC):
 @sprint-lead *draft → @product-lead *validate → @developer *develop → @quality-gate *qa-gate → @devops *push
 ```
 
-O contribuidor pode implementar a feature seguindo o [CONTRIBUTING.md](../CONTRIBUTING.md).
+O contribuidor pode implementar a feature seguindo o [CONTRIBUTING.md](../../CONTRIBUTING.md).
 
 ---
 
@@ -155,8 +155,8 @@ Contribuidores cujas ideias sao implementadas recebem credito em:
 ## Perguntas?
 
 - Abra uma Discussion na categoria **Q&A**
-- Consulte o [Troubleshooting Guide](troubleshooting.md)
+- Consulte o [Troubleshooting Guide](../troubleshooting.md)
 
 ---
 
-_Veja tambem: [CONTRIBUTING.md](../CONTRIBUTING.md) | [Guiding Principles](guiding-principles.md) | [Code of Conduct](../CODE_OF_CONDUCT.md)_
+_Veja tambem: [CONTRIBUTING.md](../../CONTRIBUTING.md) | [Guiding Principles](guiding-principles.md) | [Code of Conduct](../../CODE_OF_CONDUCT.md)_

package/docs/framework/guiding-principles.md

@@ -169,7 +169,7 @@ O SINAPSE e distribuido sob licenca MIT --- qualquer pessoa pode usar, modificar
 
 Isso garante que o framework mantem coerencia e qualidade mesmo com contribuicoes externas.
 
-> Processo de contribuicao: [CONTRIBUTING.md](../CONTRIBUTING.md)
+> Processo de contribuicao: [CONTRIBUTING.md](../../CONTRIBUTING.md)
 
 ---
 

package/docs/framework/roadmap.md

@@ -1,6 +1,6 @@
 # SINAPSE Roadmap
 
-> 🇧🇷 [Versão em Português](ROADMAP-PT.md)
+> 🇧🇷 [Versão em Português](../pt/roadmap.md)
 
 This document outlines the planned development direction for SINAPSE.
 
@@ -89,15 +89,15 @@ Have a new idea? Open an [Idea Discussion](https://github.com/caioimori/sinapse-
 
 ### 3. Write an RFC
 
-For significant features that need detailed design, [submit an RFC](/.github/RFC_TEMPLATE.md) following our structured process.
+For significant features that need detailed design, [submit an RFC](../../.github/RFC_TEMPLATE.md) following our structured process.
 
 ### 4. Contribute Directly
 
-Found something you want to implement? Check our [Contributing Guide](CONTRIBUTING.md) and [Feature Process](docs/feature-process.md).
+Found something you want to implement? Check our [Contributing Guide](../../CONTRIBUTING.md) and [Feature Process](feature-process.md).
 
 ## Changelog
 
-For what's already shipped, see [CHANGELOG.md](CHANGELOG.md).
+For what's already shipped, see [CHANGELOG.md](../../CHANGELOG.md).
 
 ## Update Process
 

package/docs/framework/source-tree.md

@@ -866,7 +866,7 @@ Runtime state is persisted in `.sinapse/`:
 
 - [Coding Standards](./coding-standards.md)
 - [Tech Stack](./tech-stack.md)
-- [ADE Architecture](../architecture/ade-architecture.md) _(planned)_
+- ADE Architecture _(planned)_
 
 ---
 

package/docs/framework/versioning-and-releases.md

@@ -1,6 +1,6 @@
 # How to Release a New Version
 
-> 🌐 **EN** | [PT](./pt/versioning-and-releases.md)
+> 🌐 **EN** | [PT](../pt/versioning-and-releases.md)
 
 ---
 

package/docs/getting-started.md

@@ -20,7 +20,7 @@ O wizard detecta seu ambiente, escolhe IDE (Claude Code ou Codex), instala os 17
 
 ```bash
 npx sinapse-ai status   # squads + agentes instalados
-npx sinapse-ai doctor   # 12 health checks
+npx sinapse-ai doctor   # 16 health checks
 ```
 
 Se algo falhar: `npx sinapse-ai doctor --fix` corrige automaticamente.

package/docs/guides/agent-selection-guide.md

@@ -282,7 +282,7 @@ Retries < 3? → @developer retry with variation
 ## Full Documentation
 
 - **[ADE Complete Guide](./ade-guide.md)** - Full tutorial for Autonomous Development Engine
-- **[Agent Responsibility Matrix](../architecture/agent-responsibility-matrix.md)** - Complete boundary definitions
+- **Agent Responsibility Matrix** - Complete boundary definitions
 
 ---
 

package/docs/guides/agents/traces/README.md

@@ -23,7 +23,7 @@ This directory contains comprehensive runtime execution trace documentation for
 | @architect | Aria (Visionary) | 21 | Direct | [architect-execution-trace.md](./architect-execution-trace.md) |
 | @developer | Dex (Builder) | 36 | Direct | [dev-execution-trace.md](./dev-execution-trace.md) |
 | @quality-gate | Quinn (Guardian) | 25 | Direct | [qa-execution-trace.md](./qa-execution-trace.md) |
-| @sinapse-orqx | Nova (Orchestrator) | 33 | Direct | [sinapse-orqx-execution-trace.md](./sinapse-orqx-execution-trace.md) |
+| @sinapse-orqx | Nova (Orchestrator) | 33 | Direct | [snps-orqx-execution-trace.md](./snps-orqx-execution-trace.md) |
 | @devops | Gage (Pipeline) | 30 | CLI Wrapper | [devops-execution-trace.md](./devops-execution-trace.md) |
 | @data-engineer | Dara (Architect) | 27 | CLI Wrapper | [data-engineer-execution-trace.md](./data-engineer-execution-trace.md) |
 | @ux-design-expert | Uma (Harmonizer) | 24 | CLI Wrapper | [ux-design-expert-execution-trace.md](./ux-design-expert-execution-trace.md) |

package/docs/guides/config-migration-guide.md

@@ -2,7 +2,7 @@
 
 Migrate from monolithic `core-config.yaml` to the layered configuration hierarchy.
 
-**ADR:** [ADR-PRO-002 — Configuration Hierarchy](../architecture/adr/adr-pro-002-configuration-hierarchy.md)
+**ADR:** ADR-PRO-002 — Configuration Hierarchy
 **Story:** PRO-4 — Core-Config Split Implementation
 
 ---

package/docs/guides/development-setup.md

@@ -1103,7 +1103,7 @@ node --inspect node_modules/.bin/jest
 ## Related Documentation
 
 - [User Guide](./user-guide.md) - End-user documentation
-- [Architecture Overview](../architecture/ARCHITECTURE-INDEX.md) - System design
+- [Architecture Overview](../pt/architecture/ARCHITECTURE-INDEX.md) - System design
 - [Contributing Squads](./contributing-squads.md) - Squad development
 - [Quality Gates Guide](./quality-gates.md) - Quality assurance
 - [MCP Global Setup](./mcp-global-setup.md) - MCP configuration

package/docs/guides/docker-mcp-setup.md

@@ -415,11 +415,11 @@ Only use docker-gateway when:
 
 ## Related Documentation
 
-- [Docker Gateway Tutorial](./guides/mcp/docker-gateway-tutorial.md)
-- [Desktop Commander MCP Guide](./guides/mcp/desktop-commander.md)
-- [MCP Global Setup Guide](./guides/mcp-global-setup.md)
-- [MCP Architecture Diagrams](./architecture/mcp-system-diagrams.md)
-- [DevOps Agent](../.sinapse-ai/development/agents/devops.md)
+- [Docker Gateway Tutorial](./mcp/docker-gateway-tutorial.md)
+- [Desktop Commander MCP Guide](./mcp/desktop-commander.md)
+- [MCP Global Setup Guide](./mcp-global-setup.md)
+- MCP Architecture Diagrams
+- [DevOps Agent](../../.sinapse-ai/development/agents/devops.md)
 
 ---
 

package/docs/guides/getting-started.md

@@ -176,7 +176,7 @@ npx sinapse-ai info
 - **Explorar os squads** — `@snps-orqx *status` mostra todos os dominios disponiveis (alias `@sinapse-orqx` ate v1.3.0)
 - **Entender os workflows** — Leia `docs/guides/workflows-overview.md`
 - **Integracao com IDEs** — Veja [IDE Integration](./ide-integration.md) para Claude Code, Gemini CLI, Codex CLI, Cursor e GitHub Copilot
-- **Contribuir** — Leia [CONTRIBUTING.md](../CONTRIBUTING.md)
+- **Contribuir** — Leia [CONTRIBUTING.md](../../CONTRIBUTING.md)
 
 ---
 

package/docs/guides/git-workflow-guide.md

@@ -1,6 +1,6 @@
 # SINAPSE Git Workflow Guide
 
-> 🌐 **EN** | [PT](./pt/git-workflow-guide.md)
+> 🌐 **EN** | [PT](../pt/git-workflow-guide.md)
 
 ---
 
@@ -978,8 +978,8 @@ strategy:
 
 ## References
 
-- **SINAPSE Validator:** [.sinapse-ai/utils/sinapse-validator.js](../.sinapse-ai/utils/sinapse-validator.js)
-- **CI Workflow:** [.github/workflows/ci.yml](../.github/workflows/ci.yml)
+- **SINAPSE Validator:** [.sinapse-ai/utils/sinapse-validator.js](../../.sinapse-ai/utils/sinapse-validator.js)
+- **CI Workflow:** [.github/workflows/ci.yml](../../.github/workflows/ci.yml)
 
 ---
 

package/docs/guides/ide-integration.md

@@ -228,9 +228,9 @@ cat ~/.claude.json  # For Claude Code
 
 ## Related Documentation
 
-- [Claude Code Guide](./pt/platforms/claude-code.md)
-- [Agent Selection Guide](./guides/agent-selection-guide.md)
-- [MCP Global Setup](./guides/mcp-global-setup.md)
+- [Claude Code Guide](../pt/platforms/claude-code.md)
+- [Agent Selection Guide](./agent-selection-guide.md)
+- [MCP Global Setup](./mcp-global-setup.md)
 
 ---
 

package/docs/guides/ide-sync-guide.md

@@ -185,5 +185,5 @@ Check that the IDE is enabled in `active_ides` section.
 
 - [Squads Overview](./squads-overview.md)
 - [Agent Reference](../agent-reference-guide.md)
-- [SINAPSE Architecture](../core-architecture.md)
+- [SINAPSE Architecture](../framework/core-architecture.md)
 

package/docs/guides/mcp/desktop-commander.md

@@ -344,9 +344,9 @@ npm install -g @anthropic/desktop-commander@latest
 
 - [Docker Gateway Tutorial](./docker-gateway-tutorial.md)
 - [MCP Global Setup Guide](../mcp-global-setup.md)
-- [Docker MCP Setup](../../docker-mcp-setup.md)
+- [Docker MCP Setup](../docker-mcp-setup.md)
 - [MCP Usage Rules](../../../.claude/rules/mcp-usage.md)
-- [Agent Tool Integration](../../architecture/agent-tool-integration-guide.md)
+- [Agent Tool Integration](../../pt/architecture/agent-tool-integration-guide.md)
 
 ---
 

package/docs/guides/mcp/docker-gateway-tutorial.md

@@ -329,7 +329,7 @@ echo "~/.docker/mcp/" >> ~/.gitignore_global
 
 - [Desktop Commander MCP Guide](./desktop-commander.md)
 - [MCP Global Setup Guide](../mcp-global-setup.md)
-- [Docker MCP Setup](../../docker-mcp-setup.md)
+- [Docker MCP Setup](../docker-mcp-setup.md)
 
 ---
 

package/docs/guides/mcp-global-setup.md

@@ -687,9 +687,9 @@ Create `.mcp.json` in project root to override global settings:
 
 - [Docker Gateway Tutorial](./mcp/docker-gateway-tutorial.md)
 - [Desktop Commander MCP Guide](./mcp/desktop-commander.md)
-- [Docker MCP Setup](../docker-mcp-setup.md)
-- [Module System Architecture](../architecture/module-system.md)
-- [MCP Architecture Diagrams](../architecture/mcp-system-diagrams.md)
+- [Docker MCP Setup](./docker-mcp-setup.md)
+- [Module System Architecture](../pt/architecture/module-system.md)
+- [MCP Architecture Diagrams](../pt/architecture/mcp-system-diagrams.md)
 
 ---