sinapse-ai 1.8.0 → 1.9.1

package/LICENSE

@@ -1,63 +1,63 @@
-MIT License
-
-Copyright (c) 2025 BMad Code, LLC (BMad Method - original work)
-Copyright (c) 2025 Synkra AIOX Inc. (AIOX Framework - intermediate work)
-Copyright (c) 2025 SinapseAI Inc. (SINAPSE Framework - derivative work)
-
-Portions of this software were derived from works distributed under the
-MIT License by the copyright holders listed above. SINAPSE is NOT
-affiliated with, endorsed by, or sanctioned by any of the above parties.
-
-Permission is hereby granted, free of charge, to any person obtaining a copy
-of this software and associated documentation files (the "Software"), to deal
-in the Software without restriction, including without limitation the rights
-to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the Software is
-furnished to do so, subject to the following conditions:
-
-The above copyright notice and this permission notice shall be included in all
-copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-SOFTWARE.
-
-TRADEMARK NOTICE:
-BMad, BMad Method, and BMad Core are trademarks of BMad Code, LLC.
-These trademarks are NOT licensed under the MIT License. See the BMad Method
-TRADEMARK.md (https://github.com/bmad-code-org/BMAD-METHOD/blob/main/TRADEMARK.md)
-for detailed guidelines on trademark usage.
-
-PERSONA SIMULATION NOTICE:
-SINAPSE includes simulated agent personas inspired by real public figures
-(authors, investors, designers, advisors, story craftspeople). These personas
-exist as educational and research tools that synthesize publicly available
-work, frameworks, and ideas associated with each figure. They are used under
-fair use for teaching, research, and creative inspiration.
-
-Each simulated persona ships with an in-file disclaimer (machine-readable
-marker `<!-- BEGIN PERSONA NOTICE -->`) stating:
-
-  * The persona is a simulation, not the real person.
-  * SINAPSE is not affiliated with, endorsed by, sponsored by, or in any way
-    officially connected to the real person.
-  * The simulated persona does not represent the views, opinions, statements,
-    or recommendations of the real person.
-  * No claim of ownership, agency, or representation of the real person is
-    asserted by SINAPSE or by any user of these personas.
-  * Use at your own discretion. Persons interacting with the simulated
-    persona acknowledge that any output is the product of language-model
-    inference over public material and should not be attributed to the real
-    person.
-
-If you are a person represented by one of these simulated personas (or their
-estate / authorized representative) and you object to the simulation, please
-open an issue at https://github.com/caioimori/sinapse-ai/issues with the
-subject "Persona Simulation Removal Request" and the persona will be removed
-or renamed in the next release. We respect requests in good faith and act
-quickly.
-
+MIT License
+
+Copyright (c) 2025 BMad Code, LLC (BMad Method - original work)
+Copyright (c) 2025 Synkra AIOX Inc. (AIOX Framework - intermediate work)
+Copyright (c) 2025 SinapseAI Inc. (SINAPSE Framework - derivative work)
+
+Portions of this software were derived from works distributed under the
+MIT License by the copyright holders listed above. SINAPSE is NOT
+affiliated with, endorsed by, or sanctioned by any of the above parties.
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+
+TRADEMARK NOTICE:
+BMad, BMad Method, and BMad Core are trademarks of BMad Code, LLC.
+These trademarks are NOT licensed under the MIT License. See the BMad Method
+TRADEMARK.md (https://github.com/bmad-code-org/BMAD-METHOD/blob/main/TRADEMARK.md)
+for detailed guidelines on trademark usage.
+
+PERSONA SIMULATION NOTICE:
+SINAPSE includes simulated agent personas inspired by real public figures
+(authors, investors, designers, advisors, story craftspeople). These personas
+exist as educational and research tools that synthesize publicly available
+work, frameworks, and ideas associated with each figure. They are used under
+fair use for teaching, research, and creative inspiration.
+
+Each simulated persona ships with an in-file disclaimer (machine-readable
+marker `<!-- BEGIN PERSONA NOTICE -->`) stating:
+
+  * The persona is a simulation, not the real person.
+  * SINAPSE is not affiliated with, endorsed by, sponsored by, or in any way
+    officially connected to the real person.
+  * The simulated persona does not represent the views, opinions, statements,
+    or recommendations of the real person.
+  * No claim of ownership, agency, or representation of the real person is
+    asserted by SINAPSE or by any user of these personas.
+  * Use at your own discretion. Persons interacting with the simulated
+    persona acknowledge that any output is the product of language-model
+    inference over public material and should not be attributed to the real
+    person.
+
+If you are a person represented by one of these simulated personas (or their
+estate / authorized representative) and you object to the simulation, please
+open an issue at https://github.com/caioimori/sinapse-ai/issues with the
+subject "Persona Simulation Removal Request" and the persona will be removed
+or renamed in the next release. We respect requests in good faith and act
+quickly.
+

package/README.en.md

@@ -19,9 +19,9 @@
 
 ## What is SINAPSE?
 
-SINAPSE is an open-source meta-framework that organizes **200 AI agents into 19 specialized squads**, operating directly in the terminal via Claude Code or Codex CLI. Each agent has a defined role, each squad masters a discipline, and the entire system is governed by a **Constitution with real enforcement** -- 19 active hooks that block violations at runtime.
+SINAPSE is an open-source meta-framework that organizes **172 AI agents into 17 specialized squads**, operating directly in the terminal via Claude Code or Codex CLI. Each agent has a defined role, each squad masters a discipline, and the entire system is governed by a **Constitution with real enforcement** -- 19 active hooks that block violations at runtime.
 
-The core concept is simple: instead of a single AI assistant trying to do everything, SINAPSE structures work into specialized teams. A branding squad handles visual identity. A cybersecurity squad handles compliance and pentesting. A copywriting squad handles persuasion and conversion. Each with its own knowledge base, workflows, and tasks -- totaling **1,237 executable tasks** ready to use.
+The core concept is simple: instead of a single AI assistant trying to do everything, SINAPSE structures work into specialized teams. A branding squad handles visual identity. A cybersecurity squad handles compliance and pentesting. A copywriting squad handles persuasion and conversion. Each with its own knowledge base, workflows, and tasks -- totaling **1,200 executable tasks** ready to use.
 
 Unlike tools that just chat with AI, SINAPSE enforces discipline. The **Documentation-First** pipeline requires a story to be created and validated before any code is written. Quality gates run automatically before merge. Unauthorized agents are blocked from pushing. All via hooks that intercept operations in real time -- not after the fact.
 
@@ -31,7 +31,7 @@ Unlike tools that just chat with AI, SINAPSE enforces discipline. The **Document
 
 Generative AI has a known problem: the more you ask of it, the worse it gets. A single assistant trying to do everything -- code, copy, branding, testing, deployment -- loses context, invents features, and suffers from context amnesia after just a few long iterations.
 
-SINAPSE solves this the way human teams solve it: **coordinated specialization**. Instead of one tired generalist, you have 189 agents in 18 squads, each with a defined role, its own knowledge base, and executable tasks. An orchestrator routes your request to whoever actually knows how to solve it -- automatically, without you needing to memorize agent names or commands.
+SINAPSE solves this the way human teams solve it: **coordinated specialization**. Instead of one tired generalist, you have 172 agents in 17 squads, each with a defined role, its own knowledge base, and executable tasks. An orchestrator routes your request to whoever actually knows how to solve it -- automatically, without you needing to memorize agent names or commands.
 
 The differential isn't just the quantity of agents. It's **real governance**: 19 active hooks intercept operations at runtime, a Constitution with 10 articles governs the framework, and 6 of those articles are NON-NEGOTIABLE -- violations are blocked before execution, not detected afterwards. **Speed with rigor, without choosing between the two.**
 
@@ -60,7 +60,7 @@ npx sinapse-ai doctor
 *help               # List available commands
 ```
 
-Done. You have 18 squads operating in your terminal.
+Done. You have 17 squads operating in your terminal.
 
 ---
 
@@ -238,32 +238,31 @@ The framework ensures no step is skipped. Each gate blocks automatically if the
 
 ---
 
-## 19 Specialized Squads
+## 17 Specialized Squads
 
 Each squad is an autonomous team with its own orchestrator, specialist agents, knowledge base, tasks, and workflows.
 
 | Squad | Domain | Agents |
 |-------|--------|--------|
 | **squad-brand** | Brand strategy, archetypes, visual audit | 15 |
-| **squad-design** | Design systems, components, tokens, UI | 15 |
-| **squad-copy** | Persuasive copywriting, headlines, conversion | 14 |
+| **squad-design** | Design systems, components, tokens, UI, art direction, premium LP | 14 |
+| **squad-copy** | Persuasive copywriting, headlines, conversion | 13 |
 | **squad-council** | Strategic advisors (Munger, Dalio, Thiel, ...) | 11 |
-| **squad-storytelling** | Narrative, scripts, story frameworks | 11 |
-| **squad-commercial** | Sales, funnel, revenue, commercial pipeline | 11 |
-| **squad-paidmedia** | Meta Ads, Google Ads, campaigns, optimization | 10 |
-| **claude-code-mastery** | Advanced Claude Code, MCP, deep integration | 10 |
+| **squad-storytelling** | Narrative, scripts, story frameworks | 10 |
+| **squad-commercial** | Sales, funnel, revenue, commercial pipeline | 10 |
+| **squad-paidmedia** | Meta Ads, Google Ads, campaigns, optimization | 9 |
 | **squad-animations** | Motion design, CSS, particles, 3D | 9 |
 | **squad-cloning** | Cognitive cloning, mind synthesis, digital twins | 9 |
-| **squad-cybersecurity** | Threat intel, pentest, compliance, LGPD | 9 |
+| **squad-cybersecurity** | Threat intel, pentest, compliance, LGPD | 8 |
 | **squad-courses** | Courses, curricula, assessments, educational launch | 8 |
-| **squad-research** | Market analysis, competitive intelligence | 8 |
-| **claude-code-mastery** | Advanced Claude Code tool mastery | 8 |
+| **squad-research** | Market analysis, competitive intelligence | 7 |
+| **claude-code-mastery** | Advanced Claude Code, MCP, deep integration | 8 |
 | **squad-content** | Editorial governance, content strategy | 7 |
 | **squad-product** | Product discovery, strategy, operations | 7 |
 | **squad-growth** | Analytics, CRO, SEO, growth hacking | 7 |
-| **squad-finance** | Budget, pricing, profitability analysis | 5 |
+| **squad-finance** | Budget, pricing, profitability analysis | 8 |
 
-**Total: 18 squads, 196 specialized agents, 1,213 tasks**
+**Total: 17 squads, 172 specialized agents, 1,200 tasks**
 
 Activate any squad via its orchestrator:
 
@@ -287,7 +286,7 @@ SINAPSE supports two IDEs with deep integrations:
 | **Claude Code** | `@agent-name` | Hooks, contextual rules, deny/allow, Chrome Brain |
 | **Codex CLI** | `/skills` or `$skill-name` | Native skills, multi-model, `codex exec` for CI/CD |
 
-Both IDEs have access to all 18 squads, 189 agents, workflows, and knowledge bases. The installer detects and configures automatically.
+Both IDEs have access to all 17 squads, 172 agents, workflows, and knowledge bases. The installer detects and configures automatically.
 
 ### Parity Table
 
@@ -445,7 +444,7 @@ See [CONTRIBUTING.md](CONTRIBUTING.md) for full details.
 npx sinapse-ai install
 ```
 
-One command. 18 squads. 189 agents. Constitutional governance. All operating directly in the terminal.
+One command. 17 squads. 172 agents. Constitutional governance. All operating directly in the terminal.
 
 **[Full documentation](docs/guides/getting-started.md)** • **[Report issue](https://github.com/caioimori/sinapse-ai/issues)** • **[Discussions](https://github.com/caioimori/sinapse-ai/discussions)**
 

package/README.md

@@ -21,9 +21,9 @@
 
 ## O que e o SINAPSE?
 
-SINAPSE e um meta-framework open source que organiza **189 agentes de IA em 18 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 e governado por uma **Constitution com enforcement real** — 13 hooks ativos que bloqueiam violacoes em tempo de execucao.
+SINAPSE e 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 e governado por uma **Constitution com enforcement real** — 13 hooks ativos que bloqueiam violacoes em tempo de execucao.
 
-O conceito central e simples: em vez de um unico 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 persuasao e conversao. Cada um com sua propria knowledge base, workflows e tasks — totalizando **1.213 tasks executaveis** prontas para uso, distribuidas pelos 18 squads.
+O conceito central e simples: em vez de um unico 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 persuasao e conversao. Cada um com sua propria knowledge base, workflows e tasks — totalizando **1.200 tasks executaveis** prontas para uso, distribuidas pelos 17 squads.
 
 Diferente de ferramentas que apenas conversam com IA, o SINAPSE impoe disciplina. O pipeline **Documentation-First** exige que uma story seja criada e validada antes de qualquer linha de codigo. Quality gates rodam automaticamente antes de merge. Agentes nao autorizados sao bloqueados de fazer push. Tudo isso via hooks que interceptam operacoes em tempo real — nao depois.
 
@@ -33,7 +33,7 @@ Diferente de ferramentas que apenas conversam com IA, o SINAPSE impoe disciplina
 
 IA generativa tem um problema conhecido: quanto mais voce pede, pior fica. Um unico assistente tentando fazer tudo — codigo, copy, branding, testes, deploy — perde contexto, inventa features e sofre de context amnesia depois de poucas iteracoes longas.
 
-SINAPSE resolve isso do jeito que times humanos resolvem: **especializacao coordenada**. Em vez de um generalista cansado, voce tem 189 agentes em 18 squads, cada um com papel definido, knowledge base propria e tasks executaveis. Um orquestrador roteia seu pedido para quem realmente sabe resolver — automaticamente, sem voce precisar decorar agent names ou comandos.
+SINAPSE resolve isso do jeito que times humanos resolvem: **especializacao coordenada**. Em vez de um generalista cansado, voce tem 172 agentes em 17 squads, cada um com papel definido, knowledge base propria e tasks executaveis. Um orquestrador roteia seu pedido para quem realmente sabe resolver — automaticamente, sem voce precisar decorar agent names ou comandos.
 
 O diferencial nao e apenas quantidade de agentes. E **governanca real**: 13 hooks ativos interceptam operacoes em tempo de execucao, uma Constitution com 10 artigos rege o framework, e 6 desses artigos sao NON-NEGOTIABLE — violacoes sao bloqueadas antes de executar, nao detectadas depois. **Velocidade com rigor, sem escolher entre os dois.**
 
@@ -67,7 +67,7 @@ Configuracao detalhada: [`docs/guides/grounding-setup.md`](docs/guides/grounding
 npx sinapse-ai install
 ```
 
-O wizard detecta seu ambiente, escolhe a IDE (Claude Code ou Codex), e instala os 18 squads automaticamente. Re-rodar o comando faz upsert idempotente — preserva suas customizacoes e so atualiza o que mudou.
+O wizard detecta seu ambiente, escolhe a IDE (Claude Code ou Codex), e instala os 17 squads automaticamente. Re-rodar o comando faz upsert idempotente — preserva suas customizacoes e so atualiza o que mudou.
 
 ### 2. Verifique
 
@@ -85,7 +85,7 @@ Se algo estiver fora do lugar, `doctor --fix` corrige automaticamente.
 *help               # Lista comandos disponiveis
 ```
 
-Pronto. Voce tem 18 squads operando no seu terminal.
+Pronto. Voce tem 17 squads operando no seu terminal.
 
 > **Nota sobre `npm install`:** A partir da v10.0.0-rc.4, o SINAPSE roda um postinstall automatico que sincroniza agents para o Claude Code, cria os diretorios de runtime (`.sinapse/handoffs/`, `.sinapse/scratchpad/`) e executa um health check rapido. Para desabilitar (CI, pipelines avancadas), defina `SINAPSE_SKIP_POSTINSTALL=1` antes do `npm install`, ou rode `npm install --ignore-scripts` (comportamento nativo do npm — nenhum postinstall eh executado). Nesse caso, rode `sinapse doctor --fix` apos a instalacao para garantir que o ambiente esteja pronto.
 
@@ -279,32 +279,31 @@ O framework garante que nenhuma etapa seja pulada. Cada gate bloqueia automatica
 
 ---
 
-## 18 Squads Especializados
+## 17 Squads Especializados
 
 Cada squad e uma equipe autonoma com orquestrador, agentes especialistas, knowledge base, tasks e workflows proprios.
 
 | Squad | Dominio | Agentes |
 |-------|---------|---------|
 | **squad-brand** | Estrategia de marca, arquetipos, auditoria visual | 15 |
-| **squad-design** | Design systems, componentes, tokens, UI | 15 |
-| **squad-copy** | Copywriting persuasivo, headlines, conversao | 14 |
+| **squad-design** | Design systems, componentes, tokens, UI, art direction, LP premium | 14 |
+| **squad-copy** | Copywriting persuasivo, headlines, conversao | 13 |
 | **squad-council** | Advisors estrategicos (Munger, Dalio, Thiel, ...) | 11 |
-| **squad-storytelling** | Narrativa, roteiros, frameworks de historia | 11 |
-| **squad-commercial** | Vendas, funil, revenue, pipeline comercial | 11 |
-| **squad-paidmedia** | Meta Ads, Google Ads, campanhas, otimizacao | 10 |
-| **claude-code-mastery** | Claude Code avancado, MCP, integracao profunda | 10 |
+| **squad-storytelling** | Narrativa, roteiros, frameworks de historia | 10 |
+| **squad-commercial** | Vendas, funil, revenue, pipeline comercial | 10 |
+| **squad-paidmedia** | Meta Ads, Google Ads, campanhas, otimizacao | 9 |
 | **squad-animations** | Motion design, CSS, particulas, 3D | 9 |
 | **squad-cloning** | Clonagem cognitiva, mind synthesis, digital twins | 9 |
-| **squad-cybersecurity** | Threat intel, pentest, compliance, LGPD | 9 |
+| **squad-cybersecurity** | Threat intel, pentest, compliance, LGPD | 8 |
 | **squad-courses** | Cursos, curriculos, assessments, launch educacional | 8 |
-| **squad-research** | Market analysis, inteligencia competitiva | 8 |
-| **claude-code-mastery** | Dominio avancado do Claude Code como ferramenta | 8 |
+| **squad-research** | Market analysis, inteligencia competitiva | 7 |
+| **claude-code-mastery** | Claude Code avancado, MCP, integracao profunda | 8 |
 | **squad-content** | Governanca editorial, estrategia de conteudo | 7 |
 | **squad-product** | Product discovery, estrategia, operacoes | 7 |
 | **squad-growth** | Analytics, CRO, SEO, growth hacking | 7 |
-| **squad-finance** | Budget, pricing, profitability analysis | 5 |
+| **squad-finance** | Budget, pricing, profitability analysis | 8 |
 
-**Total: 18 squads, 189 agentes especializados, 1.213 tasks**
+**Total: 17 squads, 172 agentes especializados, 1.200 tasks**
 
 Cada squad e ativado via seu orquestrador:
 
@@ -328,7 +327,7 @@ O SINAPSE suporta duas IDEs com integracoes profundas:
 | **Claude Code** | `@agent-name` | Hooks, rules contextuais, deny/allow, Chrome Brain |
 | **Codex CLI** | `/skills` ou `$skill-name` | Skills nativas, multi-model, `codex exec` para CI/CD |
 
-Ambas as IDEs tem acesso a todos os 18 squads, 189 agentes, workflows e knowledge bases. O installer detecta e configura automaticamente.
+Ambas as IDEs tem acesso a todos os 17 squads, 172 agentes, workflows e knowledge bases. O installer detecta e configura automaticamente.
 
 ### Tabela de Paridade
 
@@ -488,7 +487,7 @@ Veja [CONTRIBUTING.md](CONTRIBUTING.md) para detalhes completos.
 npx sinapse-ai install
 ```
 
-Um comando. 18 squads. 189 agentes. Governanca constitucional. Tudo operando direto no terminal.
+Um comando. 17 squads. 172 agentes. Governanca constitucional. Tudo operando direto no terminal.
 
 **[Documentacao completa](docs/guides/getting-started.md)** • **[Reportar issue](https://github.com/caioimori/sinapse-ai/issues)** • **[Discussions](https://github.com/caioimori/sinapse-ai/discussions)**
 

package/bin/cli.js

@@ -135,7 +135,7 @@ function runRouter() {
     }
     // eslint-disable-next-line no-fallthrough -- process.exit above terminates; Story 10.45 piggyback fix.
     case 'list':     cmdList(); break;
-    case 'status':   cmdStatus(); break;
+    case 'status':   cmdStatus({ watch: args.includes('--watch') }); break;
     case 'ideate': {
       // Wires the IdeationEngine (self-improvement analyzers) into the CLI.
       const { cmdIdeate } = require('./commands/ideate');

package/bin/commands/install.js

@@ -149,24 +149,7 @@ async function cmdInstallGlobal(opts = {}) {
   // Honors upsert + --reconfigure semantics established by Stories 10.20/10.35.
   await promptGroundingSections({ isUpsert, reconfigure });
 
-  // Pre-install summary — surface every resolved choice before any
-  // destructive action so the user can abort if something looks wrong.
-  // This is part of the install UX hardening for the GA release: even when
-  // settings are reused (silent skip path), the user still sees what the
-  // installer is about to do.
-  logger.always('');
-  logger.always(`${BOLD}──── Resumo da instalacao ────${NC}`);
-  const langLabelMap = { pt: 'Portugues', portuguese: 'Portugues', en: 'English', english: 'English' };
-  logger.always(`  Idioma:    ${langLabelMap[language] || language}`);
-  const llmLabel = Array.isArray(llmChoice) ? llmChoice.join(', ') : String(llmChoice);
-  logger.always(`  IDE:       ${llmLabel}`);
-  logger.always(`  Modo:      ${isUpsert ? 'upsert (atualizar instalacao existente)' : 'fresh install'}`);
-  logger.always(`  Destino:   ~/.sinapse/`);
-  logger.always(`${BOLD}──────────────────────────────${NC}`);
-  logger.always('');
-  logger.always(`${BOLD}Instalando Sinapse globalmente...${NC}\n`);
-
-  // Validate package — squads live in squads/ subdirectory
+  // Validate package + compute what will be installed (squads live in squads/).
   const squadsDir = path.join(ROOT, 'squads');
   const squads = getSquads(fs.existsSync(squadsDir) ? squadsDir : ROOT);
   if (squads.length === 0) {
@@ -176,6 +159,65 @@ async function cmdInstallGlobal(opts = {}) {
     process.exit(1);
   }
 
+  // ── Didactic preview + confirmation gate ─────────────────────────
+  // Surface EXACTLY what will be installed before any destructive action, and
+  // (interactive mode only) require an explicit choice before writing anything.
+  // Non-interactive / --yes / CI skips the gate and proceeds.
+  const langLabelMap = { pt: 'Portugues', portuguese: 'Portugues', en: 'English', english: 'English' };
+  const llmLabel = Array.isArray(llmChoice) ? llmChoice.join(', ') : String(llmChoice);
+  const agentTotal = squads.reduce((a, s) => a + (s.agents || 0), 0);
+  let hookCount = 0;
+  try {
+    hookCount = fs.readdirSync(path.join(ROOT, '.claude', 'hooks')).filter(f => /\.(cjs|js|py|sh)$/.test(f)).length;
+  } catch { /* hooks dir optional */ }
+
+  logger.always('');
+  logger.always(`${BOLD}──── O que será instalado ────${NC}`);
+  logger.always(`  ${BOLD}${squads.length}${NC} squads · ${BOLD}${agentTotal}${NC} agentes especializados`);
+  logger.always(`  Orquestrador master (Imperator) — chamável por ${CYAN}@sinapse${NC} / ${CYAN}@snps${NC}`);
+  if (hookCount) logger.always(`  ${hookCount} hooks de proteção + regras do framework`);
+  logger.always(`  Editor(es): ${BOLD}${llmLabel}${NC}`);
+  logger.always(`  Idioma: ${langLabelMap[language] || language}`);
+  logger.always(`  Modo: ${isUpsert ? 'atualizar instalação existente' : 'instalação nova'}`);
+  logger.always(`  Destino: ${DIM}~/.sinapse/${NC} (núcleo) + ${DIM}~/.claude/agents/${NC} (agentes)`);
+  logger.always(`${BOLD}──────────────────────────────${NC}`);
+
+  if (detectInteractiveMode()) {
+    const inquirer = require('inquirer');
+    let proceed = false;
+    while (!proceed) {
+      const { action } = await inquirer.prompt([{
+        type: 'list',
+        name: 'action',
+        message: 'Pronto para instalar:',
+        default: 'install',
+        choices: [
+          { name: 'Instalar agora', value: 'install' },
+          { name: 'Ver em detalhe o que será instalado', value: 'details' },
+          { name: 'Cancelar (nada será alterado)', value: 'cancel' },
+        ],
+      }]);
+      if (action === 'cancel') {
+        logger.always(`\n${YELLOW}Instalação cancelada. Nada foi alterado.${NC}`);
+        return;
+      }
+      if (action === 'details') {
+        logger.always('');
+        logger.always(`${BOLD}Squads e agentes:${NC}`);
+        for (const s of squads) {
+          logger.always(`  • ${s.name.replace(/^squad-/, '')} — ${s.agents} agentes`);
+        }
+        logger.always('');
+        logger.always(`${DIM}Também instala: o orquestrador master, ${hookCount} hooks de proteção, as regras do framework e os comandos /SINAPSE:agents:* .${NC}`);
+        logger.always(`${DIM}Onde: o núcleo vai para ~/.sinapse/ e os agentes chamáveis para ~/.claude/agents/. Nada fora disso é tocado.${NC}`);
+        continue;
+      }
+      proceed = true;
+    }
+  }
+
+  logger.always(`\n${BOLD}Instalando Sinapse globalmente...${NC}\n`);
+
   // Phase 1: Copy squads to ~/.sinapse/
   logger.always(`${CYAN}Phase 1:${NC} ${isUpsert ? 'Refreshing' : 'Copying'} squads to ~/.sinapse/`);
   fs.mkdirSync(SINAPSE_HOME, { recursive: true });
@@ -244,8 +286,10 @@ async function cmdInstallGlobal(opts = {}) {
     const agentsDir = path.join(SINAPSE_HOME, squad.name, 'agents');
     if (!fs.existsSync(agentsDir)) continue;
 
-    const orqxAgents = fs.readdirSync(agentsDir).filter(f => f.endsWith('.md') && f.includes('-orqx'));
-    for (const file of orqxAgents) {
+    // agent-invocation fix: generate a command/subagent for EVERY agent (not only -orqx),
+    // so every specialist is invokable via @agent and /SINAPSE:agents:agent.
+    const squadAgents = fs.readdirSync(agentsDir).filter(f => f.endsWith('.md'));
+    for (const file of squadAgents) {
       const agentId = file.replace('.md', '');
       const meta = extractAgentMeta(path.join(agentsDir, file));
       const cmdContent = generateCommandMd(agentId, meta.name, meta.icon, squad.name, squadPath, file);
@@ -269,6 +313,25 @@ async function cmdInstallGlobal(opts = {}) {
       }
     }
   }
+
+  // Master entry points (@sinapse, @snps, @sinapse-orqx, @snps-orqx) → rich Imperator stub.
+  // Overwrites the generic command file for the orqx names and adds the short aliases the
+  // user actually types. All four carry the baked-in greeting + auto-plan mandate.
+  if (fs.existsSync(sinapseMasterDest)) {
+    const masterSquadPath = `${sinapseBase}/sinapse`;
+    const masterEntryPoints = [
+      ['sinapse-orqx', 'sinapse-orqx.md'],
+      ['snps-orqx', 'snps-orqx.md'],
+      ['sinapse', 'sinapse-orqx.md'],
+      ['snps', 'snps-orqx.md'],
+    ];
+    for (const [id, personaFile] of masterEntryPoints) {
+      const personaPath = `${masterSquadPath}/agents/${personaFile}`;
+      const stub = generateMasterStub(id, personaPath, masterSquadPath, squads);
+      fs.writeFileSync(path.join(CLAUDE_COMMANDS_DIR, `${id}.md`), stub);
+      writtenAgents.add(`${id}.md`);
+    }
+  }
   logger.always(`  ${GREEN}OK${NC} ${writtenAgents.size} total command files`);
 
   // Phase 2b: Install global agents based on LLM choice
@@ -425,6 +488,29 @@ async function cmdInstallGlobal(opts = {}) {
     logger.always(`  ${DIM}Run 'npx sinapse-ai install' in your project later to complete setup${NC}`);
   }
 
+  // Phase 8b: Authoritative global-agent reconciliation. The project wizard invoked in
+  // Phase 8 re-installs a small subset (the orqx agents) to ~/.claude/agents/ in an older
+  // stale format, clobbering the rich stubs written in Phase 2b (incl. the Imperator
+  // greeting on @sinapse/@snps/-orqx). Re-copy the generated command files LAST so the
+  // rich, frontmatter'd stubs are always the final word. Idempotent.
+  try {
+    const reTargets = [];
+    if (llmChoice === 'claude-code' || llmChoice === 'both') reTargets.push(path.join(HOME, '.claude', 'agents'));
+    if (llmChoice === 'codex' || llmChoice === 'both') reTargets.push(path.join(HOME, '.codex', 'agents'));
+    if (reTargets.length && fs.existsSync(CLAUDE_COMMANDS_DIR)) {
+      const cmdFiles = fs.readdirSync(CLAUDE_COMMANDS_DIR).filter(f => f.endsWith('.md'));
+      for (const dir of reTargets) {
+        fs.mkdirSync(dir, { recursive: true });
+        for (const f of cmdFiles) {
+          fs.copyFileSync(path.join(CLAUDE_COMMANDS_DIR, f), path.join(dir, f));
+        }
+      }
+      logger.always(`  ${GREEN}OK${NC} Global agents reconciled (${cmdFiles.length} files — rich stubs authoritative)`);
+    }
+  } catch (e) {
+    logger.always(`  ${YELLOW}WARN${NC} Global agent reconciliation: ${e.message}`);
+  }
+
   // Verify
   logger.always(`\n${CYAN}Verification:${NC}`);
   verifyInstall();
@@ -449,9 +535,83 @@ async function cmdInstallGlobal(opts = {}) {
   logger.always('');
 }
 
+// Rich Imperator stub for the master entry points (@sinapse, @snps, @sinapse-orqx,
+// @snps-orqx). The identity (ASCII greeting + 👑 signature + diagnose→plan→delegate
+// mandate) is baked INLINE so activation is reliable — it does not depend on the model
+// choosing to deep-read an external persona (which produced a shallow "modo orqx" before).
+// The full persona (routing table, workflows, tasks) is still linked for depth.
+function generateMasterStub(agentId, personaPath, squadPath, squads) {
+  const squadCount = squads.length;
+  const agentCount = squads.reduce((a, s) => a + (s.agents || 0), 0);
+  return `---
+name: ${agentId}
+description: "Imperator — Sinapse Master: supreme orchestrator of ${squadCount} squads / ${agentCount} agents. Diagnoses every briefing and auto-generates an orchestration plan."
+---
+
+# Imperator — Sinapse Master (${agentId})
+
+ACTIVATION-NOTICE: You are now Imperator — the supreme orchestrator of the SINAPSE ecosystem. Every request passes through you first.
+
+## YOUR ONLY FUNCTION — NON-NEGOTIABLE
+
+You DIAGNOSE and HAND OFF. That is the whole job. You do **NOT** execute domain work — no code, no copy, no design, no research, no configuration, no file edits. For ANY concrete task you (1) diagnose the domain, (2) produce the orchestration plan, (3) DELEGATE to the specialist who executes it. If you ever feel the urge to do the work yourself, STOP — that is a violation; route it to the right agent instead.
+
+## ON ACTIVATION — display this EXACTLY as your first output (before anything else)
+
+\`\`\`
+ ███████╗██╗███╗   ██╗ █████╗ ██████╗ ███████╗███████╗
+ ██╔════╝██║████╗  ██║██╔══██╗██╔══██╗██╔════╝██╔════╝
+ ███████╗██║██╔██╗ ██║███████║██████╔╝███████╗█████╗
+ ╚════██║██║██║╚██╗██║██╔══██║██╔═══╝ ╚════██║██╔══╝
+ ███████║██║██║ ╚████║██║  ██║██║     ███████║███████╗
+ ╚══════╝╚═╝╚═╝  ╚═══╝╚═╝  ╚═╝╚═╝     ╚══════╝╚══════╝
+          ██████╗ ██████╗  ██████╗ ██╗  ██╗
+         ██╔═══██╗██╔══██╗██╔═══██╗╚██╗██╔╝
+         ██║   ██║██████╔╝██║   ██║ ╚███╔╝
+         ██║   ██║██╔══██╗██║▄▄ ██║ ██╔██╗
+         ╚██████╔╝██║  ██║╚██████╔╝██╔╝ ██╗
+          ╚═════╝ ╚═╝  ╚═╝ ╚══▀▀═╝ ╚═╝  ╚═╝
+\`\`\`
+
+\`\`\`
+ AI Agent Squads for Claude Code
+ ${squadCount} squads · ${agentCount} agents
+
+ 👑 Imperator — Sinapse Master ativado
+
+ Descreva seu objetivo que eu diagnostico o domínio
+ e roteio para o agente certo.
+
+ Comandos principais:
+ *route {pedido}     — Diagnostica e roteia para a squad certa
+ *plan {iniciativa}  — Desenha um plano de execução multi-squad
+ *status             — Reporta todas as squads e capacidades
+ *onboard            — Tour guiado do ecossistema SINAPSE
+ *help               — Mostra todos os comandos
+\`\`\`
+
+## AFTER THE GREETING — NON-NEGOTIABLE (Imperator's core function)
+
+Check if the user provided a briefing with the activation:
+- **Briefing provided** → proceed IMMEDIATELY: run the Initial State Audit → classify the request → produce an ORCHESTRATION PLAN (phases + squads + specific agents assigned + handoffs) → delegate to the specialists. NEVER ask "do you want me to plan?" — for Imperator the answer is always YES.
+- **Bare activation** → await the briefing, then apply the same flow automatically on receipt.
+
+You diagnose and DELEGATE — never execute domain work yourself. Route simple, well-defined requests directly to @specialist; complex ones to @{domain}-orqx; cross-domain ones by coordinating multiple orchestrators.
+
+## FULL OPERATING PARAMETERS
+
+For your complete persona — the routing table of all squads, the Initial State Audit, the Documentation-First gate, NSN enforcement, and every workflow and task — read and ADOPT:
+\`${personaPath}\`
+Tasks: \`${squadPath}/tasks/\` · Workflows: \`${squadPath}/workflows/\` · Manifest: \`${squadPath}/squad.yaml\`
+
+— Imperator, orchestrating SINAPSE
+`;
+}
+
 function generateCommandMd(agentId, agentName, agentIcon, squadName, squadPath, agentFile) {
   // Detecta se é orquestrador: id termina em -orqx OU é Imperator (snps-orqx/sinapse-orqx)
-  const isOrchestrator = /-orqx$/.test(agentId) || agentId === 'snps-orqx' || agentId === 'sinapse-orqx';
+  // ou um dos apelidos curtos do master (sinapse/snps).
+  const isOrchestrator = /-orqx$/.test(agentId) || agentId === 'snps-orqx' || agentId === 'sinapse-orqx' || agentId === 'sinapse' || agentId === 'snps';
 
   const step4 = isOrchestrator
     ? `4. Briefing-on-activation check (this agent is an ORCHESTRATOR):
@@ -465,7 +625,19 @@ function generateCommandMd(agentId, agentName, agentIcon, squadName, squadPath,
    - If bare activation → await briefing, then plan automatically. NEVER ask "do you want me to plan?".`
     : `6. HALT and await user input`;
 
-  return `# ${agentId}
+  // Frontmatter (name + description) is REQUIRED so the file resolves as a Claude Code
+  // subagent (@id) — without it, @id matches nothing. The same file doubles as a slash
+  // command (/SINAPSE:agents:id); extra frontmatter is harmless there.
+  const fmDesc = (isOrchestrator
+    ? `${agentName || agentId} — orchestrator for ${squadName}; diagnoses, routes to specialists and auto-generates an orchestration plan`
+    : `${agentName || agentId} — ${squadName} specialist`).replace(/["\n\r]/g, ' ').trim();
+
+  return `---
+name: ${agentId}
+description: "${fmDesc}"
+---
+
+# ${agentId}
 
 ACTIVATION-NOTICE: This command activates an agent from ${squadName}.
 

package/bin/commands/status.js

@@ -56,8 +56,21 @@ function cmdList() {
   logger.always('');
 }
 
-function cmdStatus() {
+function cmdStatus({ watch = false } = {}) {
   const logger = getLogger();
+
+  // D7 — live observability panel. `--watch` drives the animated panel from
+  // the runtime status sources (Bob dashboard + per-cwd session-cache).
+  if (watch) {
+    const { ObservabilityPanel } = require('../../.sinapse-ai/core/ui/observability-panel');
+    const handle = ObservabilityPanel.watchFromStatusFile(process.cwd(), {});
+    process.on('SIGINT', () => {
+      handle.stop();
+      process.exit(0);
+    });
+    return;
+  }
+
   header();
 
   // Check global install

package/bin/commands/uninstall.js

@@ -51,8 +51,8 @@ async function removeGitHooksConfig(projectDir = process.cwd()) {
 // Story 10.40 — Remove SINAPSE-authored orqx agents from a global agents dir.
 // Returns { removed: N } for reporting. Only touches files matching *-orqx.md
 // so we don't accidentally remove user-authored agents.
-// Audit 1 P0 (UN-1) — install writes ~200 agent files to ~/.claude/agents/ +
-// ~/.codex/agents/ but uninstall historically removed only `*-orqx.md` (~21
+// Audit 1 P0 (UN-1) — install writes ~170 agent files to ~/.claude/agents/ +
+// ~/.codex/agents/ but uninstall historically removed only `*-orqx.md` (~18
 // files). Files were left orphaned. Fix: install records every authored
 // filename in ~/.sinapse/installed-agents.json; uninstall reads that manifest
 // and removes only those files (preserving anything the user added by hand).