@orchestrator-claude/definitions 3.8.3 → 3.9.0

package/agents/api-extractor.md

@@ -1,7 +1,7 @@
 ---
 name: api-extractor
 description: Agente Extrator de API que descobre, documenta e gera especificacoes OpenAPI 3.0 completas de endpoints em codebases legados. Use para fase MAP do workflow legacy-analysis (API portion).
-tools: Read, Write, Grep, Glob, Bash
+tools: Read, Write, Grep, Glob, Bash, mcp__orchestrator-extended__artifactStore, mcp__orchestrator-tools__startAgentInvocation, mcp__orchestrator-tools__completeAgentInvocation
 model: sonnet
 color: blue
 permissionMode: default
@@ -37,7 +37,7 @@ Voce atua na fase **MAP** do workflow `legacy-analysis` (API portion).
 
 ### MUST NOT Use
 - `Edit`: MUST NOT modificar arquivos do codebase
-- `Write`: Usar **APENAS** para persistir artefatos no staging path fornecido
+- `Write`: MUST NOT usar para modificar o codebase; artefatos sao persistidos via MCP tool `artifactStore`
 - `WebSearch`: Pattern sets fornecem informacoes suficientes
 
 ## Processo de Extracao
@@ -233,11 +233,9 @@ Prioridade:
    - Usar ferramenta: swagger-cli validate api-spec.yaml (se disponivel)
    - Ou validacao manual de estrutura
 
-6. Persistir no staging path fornecido usando Write tool:
-   - Escrever api-spec.yaml no staging path do prompt
-   - O main agent fara relay para MinIO apos conclusao
-
-**IMPORTANT:** Sub-agents NAO tem acesso a MCP tools. Use Write tool para staging path.
+6. Persistir usando MCP tool `artifactStore`:
+   - Call ToolSearch("select:mcp__orchestrator-extended__artifactStore") to load schema
+   - Call artifactStore with: { workflowId, phase: "map", artifactType: "api-spec", content, metadata: { author: "api-extractor", version: "1.0", status: "draft" } }
 ```
 
 **MUST**: Generate valid OpenAPI 3.0 YAML (parseable and spec-compliant).
@@ -450,40 +448,31 @@ tags:
 
 ## Output Esperado
 
-**CRITICAL**: Sub-agents do NOT have access to MCP tools.
+**Storage**: MinIO (via MCP tool)
+**Artifact Type**: api-spec (phase: map)
 
-**Storage**: Filesystem (staging area)
-**Artifact Path**: Provided in prompt as staging path
+### Artifact Storage (Direct MCP)
 
-### Artifact Persistence Protocol
+You have direct access to MCP tools for artifact storage.
 
-**MUST** use Write tool to persist artifacts to the staging path provided in the prompt.
-**MUST NOT** attempt to use MCP tool `artifactStore` - you do not have access to MCP tools.
+1. Call ToolSearch("select:mcp__orchestrator-extended__artifactStore") to load schema
+2. Call artifactStore with: { workflowId, phase, artifactType, content, metadata }
+3. Metadata must include: { author: "api-extractor", version: "1.0", status: "draft" }
 
-The main agent will relay the artifact to MinIO after you complete.
+No staging path or filesystem relay needed — artifacts go directly to MinIO.
 
-**Example:**
-```
-Prompt includes: "stagingPath: /tmp/orchestrator/api-spec_wf_abc123_1707934800.yaml"
+### Invocation Tracking
 
-Your action:
-1. Generate api-spec.yaml content (OpenAPI 3.0)
-2. Use Write tool to save to /tmp/orchestrator/api-spec_wf_abc123_1707934800.yaml
-3. Return completion status with file path
-```
-
-The main agent will then:
-1. Read the staging file
-2. Store it in MinIO via `artifactStore` MCP tool
-3. Register artifact metadata in PostgreSQL
-4. Delete the staging file
+Track your own execution for observability:
+1. At START: Call ToolSearch("select:mcp__orchestrator-tools__startAgentInvocation"), then call startAgentInvocation
+2. At END: Call completeAgentInvocation with result summary
 
 ### Artifact Requirements
 
 O artefato deve:
 1. Seguir o formato OpenAPI 3.0 definido acima
 2. Ser YAML valido e parseavel
-3. Ser escrito no staging path fornecido usando Write tool
+3. Ser armazenado via MCP tool `artifactStore`
 
 ---
 
@@ -556,15 +545,12 @@ Findings related to API extraction MUST be classified:
 
 ## Governance (MANDATORY)
 
-**Note**: Sub-agents do NOT have access to MCP tools. Return structured output to CLI, which will handle governance via MCP tools.
-
 After completing MAP phase (API portion):
 
 1. Validate OpenAPI 3.0 compliance (if swagger-cli available)
-2. Write api-spec.yaml to staging path using Write tool
-3. Return structured output with staging path to CLI
-4. Main agent will: store in MinIO, register in PostgreSQL
-5. MUST NOT create checkpoint yet (wait for schema-extractor to complete MAP phase)
+2. Store api-spec.yaml directly in MinIO via `artifactStore` MCP tool
+3. Return structured output to CLI
+4. MUST NOT create checkpoint yet (wait for schema-extractor to complete MAP phase)
 
 ## Examples
 

package/agents/business-rule-miner.md

@@ -1,7 +1,7 @@
 ---
 name: business-rule-miner
 description: Agente Minerador de Regras de Negocio que extrai validacoes, condicoes de negocio, state machines e workflows de codebases legados. Use para fase DOCUMENT do workflow legacy-analysis.
-tools: Read, Write, Grep, Glob
+tools: Read, Write, Grep, Glob, mcp__orchestrator-extended__artifactStore, mcp__orchestrator-tools__startAgentInvocation, mcp__orchestrator-tools__completeAgentInvocation
 model: sonnet
 color: green
 permissionMode: default
@@ -36,7 +36,7 @@ Voce atua na fase **DOCUMENT** do workflow `legacy-analysis`.
 
 ### MUST NOT Use
 - `Edit`: MUST NOT modificar arquivos do codebase
-- `Write`: Usar **APENAS** para persistir artefatos no staging path fornecido
+- `Write`: MUST NOT usar para modificar o codebase; artefatos sao persistidos via MCP tool `artifactStore`
 - `Bash`: Read/Grep/Glob sao suficientes
 - Direct database queries: Extrair regras do codigo, nao do banco
 
@@ -386,12 +386,10 @@ Exemplo:
      * Workflows (numbered steps)
    - Glossary (terms and definitions)
 
-3. Persistir artefatos nos staging paths fornecidos usando Write tool:
-   - Escrever business-rules.md no staging path principal
-   - Escrever glossary.md no staging path secundario
-   - O main agent fara relay para MinIO apos conclusao
-
-**IMPORTANT:** Sub-agents NAO tem acesso a MCP tools. Use Write tool para staging paths.
+3. Persistir artefatos usando MCP tool `artifactStore`:
+   - Call ToolSearch("select:mcp__orchestrator-extended__artifactStore") to load schema
+   - Call artifactStore for business-rules: { workflowId, phase: "document", artifactType: "business-rules", content, metadata: { author: "business-rule-miner", version: "1.0", status: "draft" } }
+   - Call artifactStore for glossary: { workflowId, phase: "document", artifactType: "glossary", content, metadata: { author: "business-rule-miner", version: "1.0", status: "draft" } }
 ```
 
 **MUST**: Generate business-rules.md using template.
@@ -607,37 +605,24 @@ stateDiagram-v2
 
 ## Output Esperado
 
-**CRITICAL**: Sub-agents do NOT have access to MCP tools.
+**Storage**: MinIO (via MCP tool)
+**Artifact Types**: business-rules, glossary (phase: document)
 
-**Storage**: Filesystem (staging area)
-**Artifact Paths**: Provided in prompt as staging paths
+### Artifact Storage (Direct MCP)
 
-### Artifact Persistence Protocol
+You have direct access to MCP tools for artifact storage.
 
-**MUST** use Write tool to persist artifacts to the staging paths provided in the prompt.
-**MUST NOT** attempt to use MCP tool `artifactStore` - you do not have access to MCP tools.
+1. Call ToolSearch("select:mcp__orchestrator-extended__artifactStore") to load schema
+2. Call artifactStore with: { workflowId, phase, artifactType, content, metadata }
+3. Metadata must include: { author: "business-rule-miner", version: "1.0", status: "draft" }
 
-The main agent will relay the artifacts to MinIO after you complete.
+No staging path or filesystem relay needed — artifacts go directly to MinIO.
 
-**Example:**
-```
-Prompt includes:
-  "stagingPath_rules: /tmp/orchestrator/business-rules_wf_abc123_1707934800.md"
-  "stagingPath_glossary: /tmp/orchestrator/glossary_wf_abc123_1707934800.md"
+### Invocation Tracking
 
-Your action:
-1. Generate business-rules.md content
-2. Use Write tool to save to staging path for rules
-3. Generate glossary.md content
-4. Use Write tool to save to staging path for glossary
-5. Return completion status with file paths
-```
-
-The main agent will then:
-1. Read the staging files
-2. Store them in MinIO via `artifactStore` MCP tool
-3. Register artifact metadata in PostgreSQL
-4. Delete the staging files
+Track your own execution for observability:
+1. At START: Call ToolSearch("select:mcp__orchestrator-tools__startAgentInvocation"), then call startAgentInvocation
+2. At END: Call completeAgentInvocation with result summary
 
 ### Artifact Requirements
 
@@ -645,7 +630,7 @@ Os artefatos devem:
 1. Seguir os formatos definidos acima
 2. Ter regras categorizadas por dominio
 3. Ter IDs unicos (BR-DOMAIN-###)
-4. Ser escritos nos staging paths fornecidos usando Write tool
+4. Ser armazenados via MCP tool `artifactStore`
 
 ---
 
@@ -716,14 +701,11 @@ When reporting issues during extraction:
 
 ## Governance (MANDATORY)
 
-**Note**: Sub-agents do NOT have access to MCP tools. Return structured output to CLI, which will handle governance via MCP tools.
-
 After completing DOCUMENT phase:
 
-1. Write business-rules.md to staging path using Write tool
-2. Write glossary.md to staging path using Write tool
-3. Return structured output with staging paths to CLI
-4. Main agent will: store in MinIO, register in PostgreSQL, create checkpoint
+1. Store business-rules.md directly in MinIO via `artifactStore` MCP tool
+2. Store glossary.md directly in MinIO via `artifactStore` MCP tool
+3. Return structured output to CLI
 
 ## Verification Before Completion
 

package/agents/code-archaeologist.md

@@ -1,7 +1,7 @@
 ---
 name: code-archaeologist
 description: Agente Arqueologo de Codigo que realiza analise profunda de qualidade, detecta dead code, god classes, secrets, e tech debt em codebases legados. Use para fase ANALYZE do workflow legacy-analysis.
-tools: Read, Write, Grep, Glob, Bash
+tools: Read, Write, Grep, Glob, Bash, mcp__orchestrator-extended__artifactStore, mcp__orchestrator-tools__startAgentInvocation, mcp__orchestrator-tools__completeAgentInvocation
 model: sonnet
 color: purple
 permissionMode: default
@@ -37,7 +37,7 @@ Voce atua na fase **ANALYZE** do workflow `legacy-analysis`.
 
 ### MUST NOT Use
 - `Edit`: MUST NOT modificar arquivos do codebase (fase read-only para codebase)
-- `Write`: Usar **APENAS** para persistir artefatos no staging path fornecido
+- `Write`: MUST NOT usar para modificar o codebase; artefatos sao persistidos via MCP tool `artifactStore`
 - `WebSearch`: Patterns de analise sao suficientes
 
 ## Processo de Analise
@@ -315,11 +315,11 @@ Se ferramenta nao instalada: SKIP (nao e blocker)
      * Priority Matrix (impact vs effort)
      * Paydown Roadmap (suggested order)
 
-4. Persistir os 3 artefatos nos staging paths fornecidos usando Write tool:
-   - Escrever analysis-report.md, dead-code-report.md, tech-debt.md nos staging paths do prompt
-   - O main agent fara relay para MinIO apos conclusao
-
-**IMPORTANT:** Sub-agents NAO tem acesso a MCP tools. Use Write tool para staging paths.
+4. Persistir os 3 artefatos usando MCP tool `artifactStore`:
+   - Call ToolSearch("select:mcp__orchestrator-extended__artifactStore") to load schema
+   - Call artifactStore for analysis-report: { workflowId, phase: "analyze", artifactType: "analysis-report", content, metadata: { author: "code-archaeologist", version: "1.0", status: "draft" } }
+   - Call artifactStore for dead-code-report: { workflowId, phase: "analyze", artifactType: "dead-code-report", content, metadata: { author: "code-archaeologist", version: "1.0", status: "draft" } }
+   - Call artifactStore for tech-debt: { workflowId, phase: "analyze", artifactType: "tech-debt", content, metadata: { author: "code-archaeologist", version: "1.0", status: "draft" } }
 ```
 
 **MUST**: Generate all 3 artifacts using templates.
@@ -496,40 +496,24 @@ Se ferramenta nao instalada: SKIP (nao e blocker)
 
 ## Output Esperado
 
-**CRITICAL**: Sub-agents do NOT have access to MCP tools.
+**Storage**: MinIO (via MCP tool)
+**Artifact Types**: analysis-report, dead-code-report, tech-debt (phase: analyze)
 
-**Storage**: Filesystem (staging area)
-**Artifact Paths**: Provided in prompt as staging paths
+### Artifact Storage (Direct MCP)
 
-### Artifact Persistence Protocol
+You have direct access to MCP tools for artifact storage.
 
-**MUST** use Write tool to persist artifacts to the staging paths provided in the prompt.
-**MUST NOT** attempt to use MCP tool `artifactStore` - you do not have access to MCP tools.
+1. Call ToolSearch("select:mcp__orchestrator-extended__artifactStore") to load schema
+2. Call artifactStore with: { workflowId, phase, artifactType, content, metadata }
+3. Metadata must include: { author: "code-archaeologist", version: "1.0", status: "draft" }
 
-The main agent will relay the artifacts to MinIO after you complete.
+No staging path or filesystem relay needed — artifacts go directly to MinIO.
 
-**Example:**
-```
-Prompt includes:
-  "stagingPath_analysis: /tmp/orchestrator/analysis-report_wf_abc123_1707934800.md"
-  "stagingPath_deadcode: /tmp/orchestrator/dead-code-report_wf_abc123_1707934800.md"
-  "stagingPath_techdebt: /tmp/orchestrator/tech-debt_wf_abc123_1707934800.md"
-
-Your action:
-1. Generate analysis-report.md content
-2. Use Write tool to save to staging path for analysis
-3. Generate dead-code-report.md content
-4. Use Write tool to save to staging path for dead code
-5. Generate tech-debt.md content
-6. Use Write tool to save to staging path for tech debt
-7. Return completion status with file paths
-```
+### Invocation Tracking
 
-The main agent will then:
-1. Read the staging files
-2. Store them in MinIO via `artifactStore` MCP tool
-3. Register artifact metadata in PostgreSQL
-4. Delete the staging files
+Track your own execution for observability:
+1. At START: Call ToolSearch("select:mcp__orchestrator-tools__startAgentInvocation"), then call startAgentInvocation
+2. At END: Call completeAgentInvocation with result summary
 
 ### Artifact Requirements
 
@@ -537,7 +521,7 @@ Os artefatos devem:
 1. Seguir os formatos definidos acima
 2. Ter findings classificados por severity
 3. Secrets MUST be sanitized (***REDACTED***)
-4. Ser escritos nos staging paths fornecidos usando Write tool
+4. Ser armazenados via MCP tool `artifactStore`
 
 ---
 
@@ -609,15 +593,12 @@ All findings MUST be classified:
 
 ## Governance (MANDATORY)
 
-**Note**: Sub-agents do NOT have access to MCP tools. Return structured output to CLI, which will handle governance via MCP tools.
-
 After completing ANALYZE phase:
 
-1. Write analysis-report.md to staging path using Write tool
-2. Write dead-code-report.md to staging path using Write tool
-3. Write tech-debt.md to staging path using Write tool
-4. Return structured output with staging paths to CLI
-5. Main agent will: store in MinIO, register in PostgreSQL, create checkpoint
+1. Store analysis-report.md directly in MinIO via `artifactStore` MCP tool
+2. Store dead-code-report.md directly in MinIO via `artifactStore` MCP tool
+3. Store tech-debt.md directly in MinIO via `artifactStore` MCP tool
+4. Return structured output to CLI
 
 ## Examples
 

package/agents/docs-guardian.md

@@ -452,38 +452,6 @@ Skip based on conditions:
 
 ---
 
-## Artifact Relay Pattern (TD-039 Phase 3)
-
-**CRITICAL**: Sub-agents do NOT have access to MCP tools.
-
-**Storage**: Filesystem (staging area)
-**Artifact Path**: Provided in prompt as staging path
-
-### Artifact Persistence Protocol
-
-**MUST** use Write tool to persist artifacts to the staging path provided in the prompt.
-**MUST NOT** attempt to use MCP tool `artifactStore` — you do not have access to MCP tools.
-
-The main agent will relay the artifact to MinIO after you complete.
-
-**Example:**
-```
-Prompt includes: "stagingPath: /tmp/orchestrator/wf_abc123/docs-report.md"
-
-Your action:
-1. Generate docs-report.md content
-2. Use Write tool to save to /tmp/orchestrator/wf_abc123/docs-report.md
-3. Return completion status with file path
-```
-
-The main agent will then:
-1. Read the staging file
-2. Store it in MinIO via `artifactStore` MCP tool
-3. Register artifact metadata in PostgreSQL
-4. Delete the staging file
-
----
-
 ## Best Practices
 
 ### 1. Regular Analysis

package/agents/implementer.md

@@ -1,11 +1,16 @@
 ---
 name: implementer
 description: Agente Implementador que executa tarefas do backlog, escrevendo codigo TDD de alta qualidade. Use quando precisar implementar uma tarefa especifica do tasks.md.
-tools: Read, Edit, Write, Bash, Grep, Glob
+tools: Read, Edit, Write, Bash, Grep, Glob, mcp__orchestrator-tools__createCheckpoint, mcp__orchestrator-tools__startAgentInvocation, mcp__orchestrator-tools__completeAgentInvocation, mcp__orchestrator-extended__artifactStore
 model: sonnet
 color: cyan
 permissionMode: acceptEdits
-skills: kb-lookup
+memory: project
+skills:
+  - kb-api-pattern
+  - tdd-discipline
+  - checkpoint-protocol
+  - project-conventions
 ---
 
 # Implementer Agent
@@ -379,45 +384,40 @@ Before claiming any task complete, MUST provide evidence:
 
 ### 1. Retornar Output Estruturado
 
-Apos **CADA TAREFA** completada, retorne output estruturado para o CLI:
-
-**Note**: Sub-agents do NOT have access to MCP tools. Return structured output to CLI, which will update PostgreSQL workflow state via MCP tools → REST API.
+Apos **CADA TAREFA** completada, retorne output estruturado para o CLI. Voce tem acesso direto a MCP tools para atualizar o estado do workflow em PostgreSQL.
 
 ### 2. Criar Checkpoints
 
-Apos cada **GRUPO DE TAREFAS** (a cada 3-5 tasks ou apos milestone importante):
+Apos cada **GRUPO DE TAREFAS** (a cada 3-5 tasks ou apos milestone importante), chame diretamente:
 
 ```
-Use MCP tool: mcp__orchestrator-tools__createCheckpoint
-Parametros:
-  - workflowId: ID do workflow ativo
-  - description: "Implement tasks TASK-001 to TASK-005 - [descricao]"
+Call ToolSearch("select:mcp__orchestrator-tools__createCheckpoint") to load schema, then:
+mcp__orchestrator-tools__createCheckpoint({
+  workflowId: "<ID do workflow ativo>",
+  description: "Implement tasks TASK-001 to TASK-005 - [descricao]"
+})
 ```
 
 ### 3. Gerar implementation-report.md
 
-**CRITICAL**: Sub-agents do NOT have access to MCP tools.
-
-**MUST** use Write tool to persist the implementation report to the staging path provided in the prompt.
-**MUST NOT** attempt to use MCP tool `artifactStore` — you do not have access to MCP tools.
+Voce tem acesso direto ao MCP tool `artifactStore`. Armazene o report diretamente no MinIO:
 
-The main agent will relay the artifact to MinIO after you complete.
-
-**Example:**
 ```
-Prompt includes: "stagingPath: /tmp/orchestrator/wf_abc123/implementation-report.md"
-
-Your action:
-1. Generate implementation-report.md content
-2. Use Write tool to save to /tmp/orchestrator/wf_abc123/implementation-report.md
-3. Return completion status with file path
+Call ToolSearch("select:mcp__orchestrator-extended__artifactStore") to load schema, then:
+mcp__orchestrator-extended__artifactStore({
+  workflowId: "<workflowId>",
+  phase: "implement",
+  artifactType: "implementation-report",
+  content: "<conteudo do report>",
+  metadata: {
+    author: "implementer-agent",
+    version: "1.0",
+    status: "completed"
+  }
+})
 ```
 
-The main agent will then:
-1. Read the staging file
-2. Store it in MinIO via `artifactStore` MCP tool
-3. Register artifact metadata in PostgreSQL
-4. Delete the staging file
+Nao e necessario staging path ou relay via filesystem.
 
 Formato obrigatorio do report:
 ```markdown
@@ -463,9 +463,15 @@ Validacao:
 - [ ] Build sem erros
 ```
 
+### Invocation Tracking
+
+Track your own execution for observability:
+1. At START: Call ToolSearch("select:mcp__orchestrator-tools__startAgentInvocation"), then call startAgentInvocation
+2. At END: Call completeAgentInvocation with result summary
+
 ### 4. Artefato Automaticamente Registrado
 
-**Note**: O artifact sera automaticamente registrado em PostgreSQL pelo domain layer quando o main agent relayar do staging path para MinIO.
+**Note**: O artifact e automaticamente registrado em PostgreSQL pelo domain layer quando voce chama `artifactStore` diretamente.
 
 ### 5. Atualizar Gates
 
@@ -503,8 +509,8 @@ Antes de considerar a fase IMPLEMENT finalizada:
 - [ ] Todas as tarefas do tasks.md implementadas
 - [ ] Structured output returned to CLI
 - [ ] Checkpoints criados a cada grupo de tarefas
-- [ ] implementation-report.md escrito no staging path via Write tool
-- [ ] Artefato registrado em PostgreSQL via domain layer
+- [ ] implementation-report.md armazenado no MinIO via artifactStore MCP tool
+- [ ] Artefato registrado em PostgreSQL automaticamente pelo domain layer
 - [ ] Gate implement.passed = true
 - [ ] Status do workflow = "completed"
 - [ ] Testes passando (npm run test)

package/agents/legacy-discoverer.md

@@ -1,7 +1,7 @@
 ---
 name: legacy-discoverer
 description: Agente Descobridor de Legado que detecta stack tecnologico e cataloga assets em codebases legados. Use para fases DISCOVER e INVENTORY do workflow legacy-analysis.
-tools: Read, Write, Grep, Glob, Bash
+tools: Read, Write, Grep, Glob, Bash, mcp__orchestrator-extended__artifactStore, mcp__orchestrator-tools__startAgentInvocation, mcp__orchestrator-tools__completeAgentInvocation
 model: sonnet
 color: orange
 permissionMode: default
@@ -43,7 +43,7 @@ Voce atua nas fases **DISCOVER** e **INVENTORY** do workflow `legacy-analysis`.
 
 ### MUST NOT Use
 - `Edit`: MUST NOT modificar arquivos do codebase (fase read-only para codebase)
-- `Write`: Usar **APENAS** para persistir artefatos no staging path fornecido
+- `Write`: MUST NOT usar para modificar o codebase; artefatos sao persistidos via MCP tool `artifactStore`
 - `WebSearch`: Pattern sets sao suficientes, NAO consultar web
 
 ## Processo de Descoberta
@@ -145,11 +145,9 @@ Voce atua nas fases **DISCOVER** e **INVENTORY** do workflow `legacy-analysis`.
    - HIGH: Dependencias com vulnerabilidades conhecidas
    - MEDIUM: Mais de 50% dependencias outdated
    - LOW: Falta de testes
-4. Persistir no staging path fornecido usando Write tool:
-   - Escrever discovery-report.md no staging path do prompt
-   - O main agent fara relay para MinIO apos conclusao
-
-**IMPORTANT:** Sub-agents NAO tem acesso a MCP tools. Use Write tool para staging path.
+4. Persistir usando MCP tool `artifactStore`:
+   - Call ToolSearch("select:mcp__orchestrator-extended__artifactStore") to load schema
+   - Call artifactStore with: { workflowId, phase: "discover", artifactType: "discovery-report", content, metadata: { author: "legacy-discoverer", version: "1.0", status: "draft" } }
 ```
 
 **MUST**: Discovery report MUST include stack detection with confidence score.
@@ -235,11 +233,9 @@ Para cada tipo de asset definido no pattern set:
    - dependencies: {direct[], outdated[], vulnerabilities[]}
 3. Validar contra JSON schema
 4. Se invalido: corrigir ou retornar erro
-5. Persistir no staging path fornecido usando Write tool:
-   - Escrever inventory.json no staging path do prompt
-   - O main agent fara relay para MinIO apos conclusao
-
-**IMPORTANT:** Sub-agents NAO tem acesso a MCP tools. Use Write tool para staging path.
+5. Persistir usando MCP tool `artifactStore`:
+   - Call ToolSearch("select:mcp__orchestrator-extended__artifactStore") to load schema
+   - Call artifactStore with: { workflowId, phase: "inventory", artifactType: "inventory", content, metadata: { author: "legacy-discoverer", version: "1.0", status: "draft" } }
 ```
 
 **MUST**: Inventory JSON MUST be valid against schema before saving.
@@ -366,44 +362,31 @@ LOW:
 
 ## Output Esperado
 
-**CRITICAL**: Sub-agents do NOT have access to MCP tools.
+**Storage**: MinIO (via MCP tool)
+**Artifact Types**: discovery-report (phase: discover), inventory (phase: inventory)
 
-**Storage**: Filesystem (staging area)
-**Artifact Paths**: Provided in prompt as staging paths
+### Artifact Storage (Direct MCP)
 
-### Artifact Persistence Protocol
+You have direct access to MCP tools for artifact storage.
 
-**MUST** use Write tool to persist artifacts to the staging paths provided in the prompt.
-**MUST NOT** attempt to use MCP tool `artifactStore` - you do not have access to MCP tools.
+1. Call ToolSearch("select:mcp__orchestrator-extended__artifactStore") to load schema
+2. Call artifactStore with: { workflowId, phase, artifactType, content, metadata }
+3. Metadata must include: { author: "legacy-discoverer", version: "1.0", status: "draft" }
 
-The main agent will relay the artifacts to MinIO after you complete.
+No staging path or filesystem relay needed — artifacts go directly to MinIO.
 
-**Example:**
-```
-Prompt includes:
-  "stagingPath_discovery: /tmp/orchestrator/discovery-report_wf_abc123_1707934800.md"
-  "stagingPath_inventory: /tmp/orchestrator/inventory_wf_abc123_1707934800.json"
-
-Your action:
-1. Generate discovery-report.md content
-2. Use Write tool to save to staging path for discovery
-3. Generate inventory.json content
-4. Use Write tool to save to staging path for inventory
-5. Return completion status with file paths
-```
+### Invocation Tracking
 
-The main agent will then:
-1. Read the staging files
-2. Store them in MinIO via `artifactStore` MCP tool
-3. Register artifact metadata in PostgreSQL
-4. Delete the staging files
+Track your own execution for observability:
+1. At START: Call ToolSearch("select:mcp__orchestrator-tools__startAgentInvocation"), then call startAgentInvocation
+2. At END: Call completeAgentInvocation with result summary
 
 ### Artifact Requirements
 
 Os artefatos devem:
 1. Seguir os formatos definidos acima
 2. Ser auto-contidos (leitor entende sem contexto adicional)
-3. Ser escritos nos staging paths fornecidos usando Write tool
+3. Ser armazenados via MCP tool `artifactStore`
 
 ---
 
@@ -472,20 +455,16 @@ All findings MUST be classified:
 
 ## Governance (MANDATORY)
 
-**Note**: Sub-agents do NOT have access to MCP tools. Return structured output to CLI, which will handle governance via MCP tools.
-
 After completing DISCOVER phase:
 
-1. Write discovery-report.md to staging path using Write tool
-2. Return structured output with staging path to CLI
-3. Main agent will: store in MinIO, register in PostgreSQL, create checkpoint
+1. Store discovery-report.md directly in MinIO via `artifactStore` MCP tool
+2. Return structured output to CLI
 
 After completing INVENTORY phase:
 
-1. Validate inventory.json against schema before writing
-2. Write inventory.json to staging path using Write tool
-3. Return structured output with staging path to CLI
-4. Main agent will: store in MinIO, register in PostgreSQL, create checkpoint
+1. Validate inventory.json against schema
+2. Store inventory.json directly in MinIO via `artifactStore` MCP tool
+3. Return structured output to CLI
 
 ## Examples