@jaimevalasek/aioson 1.17.2 → 1.18.0

package/CHANGELOG.md

@@ -4,6 +4,19 @@ All notable changes to this project will be documented in this file.
 
 ## [Unreleased]
 
+## [1.18.0] - 2026-05-27
+
+### Added
+- **Agent Orchestration V2 — Phase 1: durable checkpoints, decision rationale, scoped operator memory.** Three must-have milestones shipped:
+  - **M1 — Checkpoint at gate.** `gate:approve` now writes a checkpoint JSON to `.aioson/runtime/checkpoints/gate-{A|B|C|D}-{slug}.json` after successful approval. Contains `prerequisites_snapshot` (artifact paths + mtimes), `gate_check_result`, and `decision_log`. Best-effort write (BR-AO-01: try/catch, never blocks approval). 5KB size cap with `decision_log` truncation (BR-AO-03). `workflow:heal` reads the latest checkpoint for the active feature using latest-gate-wins ordering (D > C > B > A per BR-AO-02) and injects recovery context into the healing prompt. Falls back gracefully when no checkpoint exists (EC-AO-02).
+  - **M2 — Decision rationale in handoffs.** `op:capture` appends confirmation signals to `.aioson/runtime/session-confirmations.jsonl`. `session-handoff.js` auto-collects these into `decision_rationale[]` in `last-handoff.json` (FIFO cap at 5 per BR-AO-04, only `signal=confirmation` per BR-AO-05). `dev:resume-data` passes rationale through to the context package. Accumulator cleared after each handoff.
+  - **M3 — Scoped operator memory.** `op:capture` accepts `--feature=<slug>` and `--session-id=<id>` flags (BR-AO-06: optional, NULL when omitted). Fields stored in proposal/decision markdown frontmatter and propagated on promotion. `op:list` accepts `--feature` and `--agent` filters (AND-composable per BR-AO-07). JSON output with `--feature` follows BR-AO-09 schema.
+
+### Notes
+- **QA findings resolved:** H-01 (spec correction: markdown frontmatter, not SQL ALTER TABLE) and M-02 (confidence field type documented as string `'confirmed'`). Both are spec amendments, not code changes.
+- **S1 (telemetry consumer mapping) and S2 (checkpoint lifecycle cleanup) deferred** as Should-have scope for a follow-up release.
+- **Test coverage:** 36 feature-specific tests (23 by @dev + 13 by @tester). 115/115 related tests green. Full suite regression: 2822/2816 + 1 skipped + 5 fail (all pre-existing: AC-P1-07 operator-memory, AC-ALL-101 perf flake, product kernel size, pentester text contracts, tool-invocation-hardening).
+
 ## [1.17.2] - 2026-05-22
 
 ### Security

package/README.md

@@ -559,32 +559,34 @@ aioson squad:dashboard              # real-time web monitoring panel
 
 | Agent | Role | Best for |
 |-------|------|----------|
-| `/setup` | Project onboarding & context | First step on any project |
-| `/product` | Product decisions & PRD | Feature scope, user stories |
-| `/sheldon` | Deep technical reasoning & PRD hardening | Hard engineering problems, spec review |
-| `/analyst` | Domain discovery & entity mapping | Understanding the problem space |
-| `/architect` | Project structure & technical decisions | Architecture, stack choices |
-| `/ux-ui` | UI/UX design system & component specs | Dashboards, flows, components |
-| `/pm` | Backlog & user stories | Sprint planning, task breakdown |
-| `/orchestrator` | Session protocol & parallel execution | Multi-agent coordination |
-| `/dev` | Feature implementation (any stack) | Focused dev tasks |
-| `/deyvin` / `/pair` | Pair programming & continuity | Coding — greenfield or brownfield |
-| `/qa` | Risk-first review & test generation | Quality gates before ship |
-| `/tester` | Systematic test engineering | Coverage gaps, legacy code testing |
-| `/pentester` | Adversarial security review | Security gates before release |
-| `/squad` | Parallel agent squads | Large feature sets in parallel |
-| `/genome` | Agent knowledge & learning | Adaptive squad intelligence |
-| `/committer` | Semantic commit messages | High-quality Git commits |
-| `/briefing` | Plan → structured briefing | Pre-production planning, problem framing |
-| `/copywriter` | Conversion copy & content | Marketing pages, VSL scripts |
-| `/discover` | System discovery & semantic cache | Brownfield mapping, knowledge bootstrap |
-| `/neo` | Onboarding & next steps | "Where do I start?" guidance |
-| `/orache` | Market research & competitors | External data gathering |
-| `/profiler-researcher` | Persona research & profiling | DNA mental research |
-| `/profiler-enricher` | Profile enrichment | Cognitive analysis |
-| `/profiler-forge` | Profile generation | Genome 3.0 advisor creation |
-| `/site-forge` | Site cloning & design extraction | Clone, harvest, blend, or forge skills from any URL |
-| `/validator` | Deliverable validation | Pre-gate verification |
+| `/aioson:agent:setup` | Project onboarding & context | First step on any project |
+| `/aioson:agent:product` | Product decisions & PRD | Feature scope, user stories |
+| `/aioson:agent:sheldon` | Deep technical reasoning & PRD hardening | Hard engineering problems, spec review |
+| `/aioson:agent:analyst` | Domain discovery & entity mapping | Understanding the problem space |
+| `/aioson:agent:architect` | Project structure & technical decisions | Architecture, stack choices |
+| `/aioson:agent:ux-ui` | UI/UX design system & component specs | Dashboards, flows, components |
+| `/aioson:agent:pm` | Backlog & user stories | Sprint planning, task breakdown |
+| `/aioson:agent:orchestrator` | Session protocol & parallel execution | Multi-agent coordination |
+| `/aioson:agent:dev` | Feature implementation (any stack) | Focused dev tasks |
+| `/aioson:agent:deyvin` / `/aioson:agent:pair` | Pair programming & continuity | Coding — greenfield or brownfield |
+| `/aioson:agent:qa` | Risk-first review & test generation | Quality gates before ship |
+| `/aioson:agent:tester` | Systematic test engineering | Coverage gaps, legacy code testing |
+| `/aioson:agent:pentester` | Adversarial security review | Security gates before release |
+| `/aioson:agent:squad` | Parallel agent squads | Large feature sets in parallel |
+| `/aioson:agent:genome` | Agent knowledge & learning | Adaptive squad intelligence |
+| `/aioson:agent:committer` | Semantic commit messages | High-quality Git commits |
+| `/aioson:agent:briefing` | Plan → structured briefing | Pre-production planning, problem framing |
+| `/aioson:agent:copywriter` | Conversion copy & content | Marketing pages, VSL scripts |
+| `/aioson:agent:discover` | System discovery & semantic cache | Brownfield mapping, knowledge bootstrap |
+| `/aioson:agent:neo` | Onboarding & next steps | "Where do I start?" guidance |
+| `/aioson:agent:orache` | Market research & competitors | External data gathering |
+| `/aioson:agent:profiler-researcher` | Persona research & profiling | DNA mental research |
+| `/aioson:agent:profiler-enricher` | Profile enrichment | Cognitive analysis |
+| `/aioson:agent:profiler-forge` | Profile generation | Genome 3.0 advisor creation |
+| `/aioson:agent:site-forge` | Site cloning & design extraction | Clone, harvest, blend, or forge skills from any URL |
+| `/aioson:agent:design-hybrid-forge` | Hybrid design system generation | Merge two visual parents into one |
+| `/aioson:agent:discovery-design-doc` | Discovery & design doc generation | Living design doc bridging discovery to implementation |
+| `/aioson:agent:validator` | Deliverable validation | Pre-gate verification |
 
 ---
 
@@ -673,6 +675,9 @@ Package, publish, and distribute squads, genomes, skills, and full systems throu
 aioson system:package .
 aioson system:publish . --name=my-stack
 
+# Build mode: compile TS/JSX, obfuscate JS (terser), package as ZIP
+aioson system:publish . --name=my-stack --build
+
 # Publish squads, genomes, and skills
 aioson squad:publish . --squad=marketing --version=1.0.0
 aioson genome:publish . --slug=fintech --version=1.0.0
@@ -924,36 +929,65 @@ aioson scan:project --json
 ## Documentation
 
 **CLI & commands**
-- [CLI reference](docs/en/cli-reference.md) — full docs for every command
-- [JSON schemas](docs/en/json-schemas.md) — `--json` output contracts
+- [CLI reference](docs/en/5-reference/cli-reference.md) — full docs for every command
+- [JSON schemas](docs/en/5-reference/json-schemas.md) — `--json` output contracts
 
 **Feature guides**
-- [Parallel orchestration](docs/en/parallel.md)
-- [MCP guide](docs/en/mcp.md)
-- [Browser QA guide](docs/en/qa-browser.md)
-- [Web3 guide](docs/en/web3.md)
-- [i18n guide](docs/en/i18n.md)
-- [Squad Dashboard](docs/en/squad-dashboard.md)
+- [Parallel orchestration](docs/en/5-reference/parallel.md)
+- [MCP guide](docs/en/5-reference/mcp.md)
+- [Browser QA guide](docs/en/5-reference/qa-browser.md)
+- [Web3 guide](docs/en/5-reference/web3.md)
+- [i18n guide](docs/en/5-reference/i18n.md)
+- [Squad Dashboard](docs/en/5-reference/squad-dashboard.md)
 
 **Portuguese guides**
-- [Início rápido](docs/pt/inicio-rapido.md)
+
+*Entender*
+- [O que é AIOSON](docs/pt/1-entender/o-que-e-aioson.md)
+- [Por que existe](docs/pt/1-entender/por-que-existe.md)
+- [Mapa do ecossistema](docs/pt/1-entender/mapa-do-ecossistema.md)
+- [Glossário](docs/pt/1-entender/glossario.md)
+
+*Começar*
+- [Primeiro projeto](docs/pt/2-comecar/primeiro-projeto.md)
+- [Projeto existente](docs/pt/2-comecar/projeto-existente.md)
+- [Decisões iniciais](docs/pt/2-comecar/decisoes-iniciais.md)
+
+*Receitas*
+- [App SaaS do zero](docs/pt/3-receitas/app-saas-do-zero.md)
+- [Feature completa com Sheldon](docs/pt/3-receitas/feature-completa-com-sheldon.md)
+- [Da ideia ao PRD via briefing](docs/pt/3-receitas/da-ideia-ao-prd-via-briefing.md)
+- [Landing page](docs/pt/3-receitas/landing-page.md)
+- [Continuidade entre sessões](docs/pt/3-receitas/continuidade-entre-sessoes.md)
+- [Integração em codebase grande](docs/pt/3-receitas/integracao-em-codebase-grande.md)
+- [Refatoração grande](docs/pt/3-receitas/refatoracao-grande.md)
+- [Auditoria de segurança](docs/pt/3-receitas/auditoria-seguranca.md)
+- [Clonar design de site](docs/pt/3-receitas/clonar-design-de-site.md)
+- [Publicar no AIOSON Store](docs/pt/3-receitas/publicar-no-aioson-com.md)
+
+*Agentes*
 - [Guia de agentes](docs/pt/agentes.md)
-- [Squad e Genome](docs/pt/squad-genome.md)
-- [Cenários de uso](docs/pt/cenarios.md)
-- [Clientes AI](docs/pt/clientes-ai.md)
-- [Runtime observability](docs/pt/runtime-observability.md)
-- [Recuperação de sessão](docs/pt/recuperacao-de-sessao.md)
-- [Monitor de contexto](docs/pt/monitor-de-contexto.md)
-- [Busca de contexto](docs/pt/busca-de-contexto.md)
-- [Cache de contexto](docs/pt/cache-de-contexto.md)
-- [Sandbox executor](docs/pt/sandbox.md)
-- [Agent sharding](docs/pt/agent-sharding.md)
-- [Guia do engenheiro](docs/pt/guia-engineer.md)
-- [Squad Dashboard (PT)](docs/pt/squad-dashboard.md)
-- [SDD automation scripts](docs/pt/sdd-automation-scripts.md)
-- [Motor hardening](docs/pt/motor-hardening.md)
-- [Design-docs governance](docs/pt/design-docs-governance.md)
-- [Runner system](docs/pt/runner-system.md)
+- [Referência por agente](docs/pt/4-agentes/README.md)
+
+*Referência*
+- [Clientes AI](docs/pt/5-referencia/clientes-ai.md)
+- [SDD framework](docs/pt/5-referencia/sdd-framework.md)
+- [Motor hardening](docs/pt/5-referencia/motor-hardening.md)
+- [Runtime observability](docs/pt/5-referencia/runtime-observability.md)
+- [Live sessions](docs/pt/5-referencia/live-sessions.md)
+- [Memória e contexto](docs/pt/5-referencia/memoria-e-contexto.md)
+- [Feature dossier](docs/pt/5-referencia/feature-dossier.md)
+- [Agent sharding](docs/pt/5-referencia/agent-sharding.md)
+- [Skills](docs/pt/5-referencia/skills.md)
+- [Sandbox executor](docs/pt/5-referencia/sandbox.md)
+- [Squad Dashboard (PT)](docs/pt/5-referencia/squad-dashboard.md)
+- [Design-docs governance](docs/pt/5-referencia/design-docs-governance.md)
+- [Runner system](docs/pt/5-referencia/runner-system.md)
+- [Secure by default](docs/pt/5-referencia/secure-by-default.md)
+
+*Living Memory*
+- [Memória viva](docs/pt/living-memory/memoria-viva.md)
+- [Autonomy contract](docs/pt/living-memory/autonomy-contract.md)
 
 ---
 

package/docs/en/3-recipes/full-feature-with-sheldon.md

@@ -10,7 +10,7 @@
 ## The trail in one line
 
 ```
-@product → @sheldon → @analyst → @architect → @ux-ui → @pm → @orchestrator → @dev
+/aioson:agent:product → @sheldon → @analyst → @architect → @ux-ui → @pm → @orchestrator → @dev
                                                               │
                                               (@qa ↔ @dev)  ← autonomous loop
                                                               │

package/docs/en/5-reference/cli-reference.md

@@ -225,13 +225,13 @@ aioson agents --json
 **Output example:**
 ```
 Agents (locale: en):
-@setup (setup)
+/aioson:agent:setup (setup)
   Path: .aioson/locales/en/agents/setup.md
   Active: .aioson/agents/setup.md
   Depends on: none
   Output: .aioson/context/project.context.md
 
-@product (product)
+/aioson:agent:product (product)
   Path: .aioson/locales/en/agents/product.md
   Active: .aioson/agents/product.md
   Depends on: none
@@ -261,7 +261,7 @@ aioson agent:prompt dev --tool=gemini --json
 - `--tool=codex|claude|gemini|opencode` — formats the prompt for the target CLI. Default: `codex`.
 - `--json` — returns structured JSON with the prompt string.
 
-**When to use:** if you're using an AI CLI that doesn't support `/setup` slash commands, run this to get the exact text to paste into the chat.
+**When to use:** if you're using an AI CLI that doesn't support `/aioson:agent:setup` slash commands, run this to get the exact text to paste into the chat.
 
 ```bash
 # Copy the prompt for @analyst in Gemini
@@ -310,7 +310,7 @@ Notes:
 Once the project is set up, each new feature follows a shorter sequence — no `@setup` required:
 
 ```
-@product → @analyst → @dev → @qa
+/aioson:agent:product → @analyst → @dev → @qa
 ```
 
 `@product` creates a feature-scoped `prd-{slug}.md` and registers the feature in `features.md`. `@analyst` produces `requirements-{slug}.md` and `spec-{slug}.md`. `@dev` reads the feature spec. `@qa` closes the feature by running `feature:close --verdict=PASS`, which updates `spec-{slug}.md` with a QA sign-off, marks it `done` in `features.md`, and automatically archives all feature artefacts to `.aioson/context/done/{slug}/`.

package/docs/en/5-reference/qa-browser.md

@@ -259,14 +259,14 @@ Two terminals, running in parallel:
 
 ```bash
 # Terminal 1 — aioson agents
-@setup → @product → @analyst → @architect → @dev
+/aioson:agent:setup → @product → @analyst → @architect → @dev
 
 # Terminal 2 — browser QA (while app is running)
 aioson qa:init --url=http://localhost:3000
 aioson qa:run
 
 # Terminal 1 — continues with merged findings
-@qa   # reads aios-qa-report.md automatically
+/aioson:agent:qa   # reads aios-qa-report.md automatically
 ```
 
 ---

package/docs/en/README.md

@@ -83,7 +83,7 @@ npx @jaimevalasek/aioson init my-project
 cd my-project
 
 # 3. Open your AI client (Claude Code, Codex, Gemini, OpenCode) and type:
-@setup
+/aioson:agent:setup
 ```
 
 From there, the agents guide you. Details in [First project from scratch](./2-start/first-project.md).

package/docs/en/deyvin-subtask-scout/how-to-use.md

@@ -95,7 +95,7 @@ Output:
 }
 ```
 
-Scout persists with `status: "error"`. `@deyvin` informs the user and offers manual handoff to `/architect` if needed.
+Scout persists with `status: "error"`. `@deyvin` informs the user and offers manual handoff to `/aioson:agent:architect` if needed.
 
 ---
 
@@ -121,7 +121,7 @@ Output:
 }
 ```
 
-`@deyvin` surfaces the message and asks the user how to proceed: usually handoff to `/architect` if scouts keep multiplying.
+`@deyvin` surfaces the message and asks the user how to proceed: usually handoff to `/aioson:agent:architect` if scouts keep multiplying.
 
 To increase the limit in the project:
 

package/docs/en/deyvin-subtask-scout/sub-task-scout.md

@@ -9,7 +9,7 @@
 Without the scout, `@deyvin` had two bad options when a question required inspecting many files:
 
 1. **Read everything inline** — burns ≥10k tokens in the parent context, pollutes the agent's working memory, and forces the next turn to compete with stale survey content.
-2. **Hand off to `/architect` or pause** — overshoots the actual need (most surveys don't require architectural decisions) and breaks the conversation flow with a full agent switch.
+2. **Hand off to `/aioson:agent:architect` or pause** — overshoots the actual need (most surveys don't require architectural decisions) and breaks the conversation flow with a full agent switch.
 
 The scout resolves this: the parent agent keeps its context lean and receives a structured report of ~500 tokens instead of the raw files.
 
@@ -31,7 +31,7 @@ aioson scout:prep --question="..." --scope-paths="a.js,b.js" ...
   └─ returns { id, prompt, output_path, cap_remaining }
          │
          ▼
-@deyvin calls harness.sub-agent(prompt)
+/aioson:agent:deyvin calls harness.sub-agent(prompt)
   └─ sub-agent runs in ISOLATED context
   └─ allowed tools: [Read, Grep]
   └─ disallowed tools: [Bash, Edit, Write]   ← Nautilus pattern
@@ -50,7 +50,7 @@ aioson scout:commit --input=<output_path>
   └─ emits telemetry action=committed
          │
          ▼
-@deyvin reads findings, confidence, recommendation
+/aioson:agent:deyvin reads findings, confidence, recommendation
   └─ folds into user-facing reply
   └─ parent context grew ~500 tokens (just the report)
      vs inline survey: ~10k+ tokens (raw file contents)

package/docs/en/deyvin-subtask-scout/troubleshooting.md

@@ -81,7 +81,7 @@ You have 1 retry (`max_retries_on_malformed_json=1`). If the second attempt also
 **What to do:**
 
 1. The scout was persisted with `status: "error"` at `.aioson/runtime/scouts/{id}.json`.
-2. `@deyvin` informs the user and offers handoff to `/architect` or a direct best-effort answer.
+2. `@deyvin` informs the user and offers handoff to `/aioson:agent:architect` or a direct best-effort answer.
 3. To increase retries: `{ "max_retries_on_malformed_json": 2 }` in config (rarely needed — if it happens frequently, the prompt template needs tightening).
 
 ---

package/docs/pt/3-receitas/publicar-no-aioson-com.md

@@ -189,6 +189,23 @@ O `package.json` do pacote é diferente do seu `package.json` de Node.js. É o m
 
 ---
 
+## Build antes de publicar (v1.17.2+)
+
+Se o seu pacote contém TypeScript, JSX ou código que precisa ser compilado, use a flag `--build`:
+
+```bash
+npx @jaimevalasek/aioson system:publish --slug=legal-compliance --type=squad --build
+```
+
+O `--build`:
+1. Compila TS/JSX antes de empacotar
+2. Aplica ofuscação JS via terser (minificação + mangling)
+3. Gera o pacote final em formato ZIP
+
+Útil para proteger lógica proprietária em squads e skills distribuídos via aioson.com.
+
+---
+
 ## Variações
 
 | Situação | Ajuste |

package/docs/pt/5-referencia/comandos-cli.md

@@ -179,7 +179,7 @@ A nova versão da Store permite empacotar, distribuir e instalar não só agente
 | Comando | O que faz | Quando usar |
 |---|---|---|
 | `system:package` | Lê o `system.json` e empacota o projeto local em `.aioson/system-packages` | Quando quiser testar o empacotamento completo do seu sistema antes de submetê-lo |
-| `system:publish` | Empacota e publica seu sistema/boilerplate na AIOSON Store | Quando quiser distribuir uma base arquitetural inteira para que outros comecem projetos rapidamente |
+| `system:publish` | Empacota e publica seu sistema/boilerplate na AIOSON Store. Com `--build`, compila TS/JSX e aplica ofuscação JS (terser). Pacote final em ZIP. | Quando quiser distribuir uma base arquitetural inteira para que outros comecem projetos rapidamente |
 | `system:list` | Lista os sistemas disponíveis localmente ou na nuvem | Para descobrir boilerplates e sistemas base que podem ser instalados |
 | `system:install` | Baixa e inicializa um sistema completo a partir da Store | Para dar kickstart num projeto novo a partir de um `system` já configurado com squads e arquitetura |
 | `squad:list` | Lista squads instaladas localmente ou remotamente na Store `--remote` | Para descobrir e inspecionar quais squads estão ativas ou disponíveis na nuvem |
@@ -488,7 +488,7 @@ aioson agents . --lang=pt-BR
 aioson agent:prompt architect . --tool=codex
 ```
 
-Use `agents` para ver quem existe e `agent:prompt` quando o cliente de IA nao entende `/setup`, `@dev` ou slash commands, ou quando voce quer um handoff direto rastreado no runtime antes de continuar em outro cliente.
+Use `agents` para ver quem existe e `agent:prompt` quando o cliente de IA nao entende `/aioson:agent:setup`, `@dev` ou slash commands, ou quando voce quer um handoff direto rastreado no runtime antes de continuar em outro cliente.
 
 ### 10. Validar agentes e pacote antes de release
 

package/docs/pt/5-referencia/inteligencia-adaptativa.md

@@ -64,7 +64,7 @@ aioson learning:apply . --file=.aioson/evolution/pending-2026-03-30T12-00-00.jso
 
 ### Exemplos práticos
 
-**Cenário:** durante 3 sessões com o agente `/dev`, você corrigiu o mesmo erro — o agente tentava criar arquivos `.ts` num projeto que usa apenas `.js`.
+**Cenário:** durante 3 sessões com o agente `/aioson:agent:dev`, você corrigiu o mesmo erro — o agente tentava criar arquivos `.ts` num projeto que usa apenas `.js`.
 
 O agente registrou um learning `correction` a cada sessão. Com frequência 3:
 
@@ -164,7 +164,7 @@ aioson tool:unregister . --name=run_tests
 
 ### Exemplos práticos
 
-**Cenário:** agente `/qa` precisa checar se o servidor local está no ar antes de rodar testes. Em vez de escrever o curl toda vez:
+**Cenário:** agente `/aioson:agent:qa` precisa checar se o servidor local está no ar antes de rodar testes. Em vez de escrever o curl toda vez:
 
 ```bash
 aioson tool:register . \
@@ -180,7 +180,7 @@ aioson tool:call . --name=check_health --input='{}'
 # OK
 ```
 
-**Cenário:** agente `/dev` cria um script mais complexo para validar migrações:
+**Cenário:** agente `/aioson:agent:dev` cria um script mais complexo para validar migrações:
 
 ```bash
 # Agente cria o arquivo

package/docs/pt/5-referencia/skills.md

@@ -202,7 +202,7 @@ Se você quer o fluxo completo em um só lugar, leia a página dedicada: [design
 
 Fluxo padrão:
 
-1. O usuário ativa `@design-hybrid-forge` no Codex ou `/design-hybrid-forge` no Claude
+1. O usuário ativa `@design-hybrid-forge` no Codex ou `/aioson:agent:design-hybrid-forge` no Claude
 2. Se quiser um overlay visual mais ousado, o usuário pode rodar `aioson design-hybrid:options` para gerar `.aioson/context/design-variation-preset.md`
 3. O comando usa `conversation_language` do projeto quando existir; `--locale` funciona como override manual
 4. O agente pede as 2 skills primárias, até 2 modificadores opcionais e lê o preset visual quando ele existir

package/docs/pt/5-referencia/web3.md

@@ -240,7 +240,7 @@ O @dev segue estas regras para contratos:
 O @qa em projetos Web3 vai além de testes funcionais:
 
 ```
-/qa
+/aioson:agent:qa
 
 Audite os contratos para:
 - Reentrancy (todas as funções externas que modificam estado)
@@ -288,7 +288,7 @@ Após instalar, você tem acesso a skills estáticas de referência:
 
 Para usar uma skill, referencie no seu AI IDE:
 ```
-/dev Usando a skill web3-security-checklist, audite nosso contrato.
+/aioson:agent:dev Usando a skill web3-security-checklist, audite nosso contrato.
 ```
 
 ---
@@ -762,7 +762,7 @@ async function takeDown(contract: BBoardAPI) {
 #### @qa audita:
 
 ```
-/qa
+/aioson:agent:qa
 
 Revise o bboard.compact para:
 - Replay attack: sequence.increment() executado corretamente no takeDown?

package/docs/pt/README.md

@@ -83,7 +83,7 @@ npx @jaimevalasek/aioson init meu-projeto
 cd meu-projeto
 
 # 3. Abra seu cliente AI (Claude Code, Codex, Gemini, OpenCode) e digite:
-@setup
+/aioson:agent:setup
 ```
 
 A partir daí, os agentes guiam você. Detalhes em [Primeiro projeto do zero](./2-comecar/primeiro-projeto.md).

package/docs/pt/_arquivo/README.md

@@ -65,7 +65,7 @@ npx @jaimevalasek/aioson setup:context
 npx @jaimevalasek/aioson doctor
 ```
 
-Depois, abra seu AI IDE e digite `/setup` para começar.
+Depois, abra seu AI IDE e digite `/aioson:agent:setup` para começar.
 
 Fluxo recomendado:
 - `@setup` primeiro