@atlashub/smartstack-cli 3.39.0 → 3.40.0

package/templates/skills/business-analyse/steps/step-05b-deploy.md

@@ -1,522 +1,522 @@
----
-name: step-05b-deploy
-description: Generate prd.json, deploy artifacts, interactive HTML
-model: sonnet
-next_step: steps/step-05c-ralph-readiness.md
----
-
-> **Context files:** _shared.md
-
-# Step 5b: Deploy Artifacts & Interactive HTML
-
-## MANDATORY EXECUTION RULES
-
-- **ALWAYS** verify all module handoffs are complete before generating artifacts
-- **ALWAYS** derive prd.json from feature.json (NEVER independently)
-- **NEVER** invent entities/FRs/BRs not in feature.json
-- **ALWAYS** update BA manifest at docs/index.json
-- **ALWAYS** deploy ba-interactive.html PRE-POPULATED with all data
-- **ALWAYS** follow references/html-data-mapping.md EXACTLY for FEATURE_DATA and EMBEDDED_ARTIFACTS mapping
-- **FEATURE_DATA.moduleSpecs MUST be populated for EACH module** — empty moduleSpecs = BROKEN specification tabs
-- **FEATURE_DATA.cadrage.scope must use HTML keys** (vital/important/optional/excluded) NOT feature.json keys (mustHave/shouldHave/couldHave/outOfScope)
-- **FEATURE_DATA.metadata.analysisMode must be set** from master.metadata.analysisMode (always "interactive")
-- **NEVER leave FEATURE_DATA or EMBEDDED_ARTIFACTS empty or partial**
-
-## YOUR TASK
-
-Generate deployment artifacts (prd.json, progress.txt, BA manifest) and deploy the interactive HTML document pre-populated with all analysis data.
-
----
-
-## EXECUTION SEQUENCE
-
-### 0. Pre-flight Verification
-
-Before generating ANY artifact, verify step-05a completed successfully:
-
-```
-FOR each module in modules[]:
-  Read module feature.json
-  IF module.handoff === {} OR module.status !== "handed-off":
-    → BLOCKING ERROR: Module {module.code} has no handoff data
-    → Return to step-05a-handoff.md
-```
-
-IF all modules have valid handoff → proceed.
-
----
-
-### 1. Generate prd.json (PROGRAMMATIC)
-
-> **RULE:** prd.json is extracted by CLI code, **NEVER** generated by LLM.
-> The `ss derive-prd` command performs a deterministic data transformation from feature.json.
-
-**For each module:**
-
-```
-Execute: ss derive-prd --feature {moduleFeaturePath} --output .ralph/prd-{moduleCode}.json
-```
-
-**For consolidated view (multi-module, optional):**
-
-```
-Execute: ss derive-prd --application {masterFeaturePath}
-→ Generates .ralph/prd-{moduleCode}.json for each module
-```
-
-**Verification:** After execution, read the generated prd.json and display summary:
-
-```
-prd.json generated for module {moduleCode}:
-  - Use cases: {count}
-  - Functional requirements: {count}
-  - Business rules: {count}
-  - API endpoints: {count}
-  - Sections: {count}
-  - Files to create: {count}
-  - BR-to-code mappings: {count}
-```
-
-**Key guarantees:**
-- Source MUST reference feature.json path (traceability)
-- All data is EXACT COPY from feature.json (no transformation, no invention)
-- prd.json $version: "3.0.0" (unified format with pre-computed tasks)
-- source.type: "ba-handoff-programmatic"
-
-**POST-CHECK (BLOCKING — DO NOT SKIP):**
-
-After `ss derive-prd` execution, read the generated prd-{moduleCode}.json and verify structural integrity:
-
-```javascript
-const prd = readJSON(`.ralph/prd-${moduleCode}.json`);
-const featureHandoff = moduleFeature.handoff.filesToCreate;
-
-// 1. Verify structure: filesToCreate MUST be under implementation (NOT root level)
-if (prd.filesToCreate && !prd.implementation?.filesToCreate) {
-  BLOCKING_ERROR("prd.json has filesToCreate at ROOT level — this is NOT from ss derive-prd");
-  BLOCKING_ERROR("Re-run: ss derive-prd --feature {path} --output .ralph/prd-{moduleCode}.json");
-  STOP;
-}
-
-// 2. Verify ALL 7 categories present and match feature.json
-const categories = ['domain', 'application', 'infrastructure', 'api', 'frontend', 'seedData', 'tests'];
-for (const cat of categories) {
-  const prdCount = prd.implementation.filesToCreate[cat]?.length ?? 0;
-  const featureCount = featureHandoff[cat]?.length ?? 0;
-  if (prdCount !== featureCount) {
-    BLOCKING_ERROR(`${cat}: prd has ${prdCount} files but feature.json has ${featureCount}`);
-  }
-}
-```
-
-Display verification table:
-
-```
-POST-CHECK: prd-{moduleCode}.json integrity
-| Category       | feature.json | prd.json | Match |
-|----------------|-------------|----------|-------|
-| domain         | {n}         | {n}      | OK/FAIL |
-| application    | {n}         | {n}      | OK/FAIL |
-| infrastructure | {n}         | {n}      | OK/FAIL |
-| api            | {n}         | {n}      | OK/FAIL |
-| frontend       | {n}         | {n}      | OK/FAIL |
-| seedData       | {n}         | {n}      | OK/FAIL |
-| tests          | {n}         | {n}      | OK/FAIL |
-```
-
-IF ANY category shows FAIL → **STOP AND RE-RUN `ss derive-prd`**. DO NOT proceed with incomplete PRD.
-
----
-
-### 2. Initialize Progress Tracker
-
-> **Template:** Read `templates/tpl-progress.md` for the complete progress tracker template structure.
-> Populate the template with module-specific data from feature.json handoff sections.
-
-**Progress Tracker Rules:**
-- One section per module, in topological order (dependencies first)
-- CORE SeedData ALWAYS 5 entries (mandatory)
-- Business SeedData varies by module
-- Hierarchical task structure (module -> layer -> tasks)
-- Each task is independent, assignable checkbox
-- Effort estimate per module (simple/medium/complex)
-- Summary with totals across all modules
-- Cross-module tasks only if multi-module
-
----
-
-### 3. Update BA Manifest (MANDATORY)
-
-> **The BA manifest enables the SmartStack web app to discover and display all available business analyses.**
-> It is a JSON index file at `docs/index.json` that lists all feature.json files.
-
-**Path:** `docs/index.json` (project root relative)
-
-**Schema:**
-```json
-{
-  "version": "1.0",
-  "updatedAt": "{ISO timestamp}",
-  "analyses": [
-    {
-      "appCode": "{app_code}",
-      "appName": "{application_name}",
-      "moduleCode": null | "{module_code}",
-      "moduleName": "{module_name}",
-      "version": "{version}",
-      "status": "handed-off",
-      "featureDescription": "{feature_description}",
-      "path": "{app_code}/business-analyse/v{version}/feature.json",
-      "updatedAt": "{ISO timestamp}"
-    }
-  ]
-}
-```
-
-> Note: The "path" field uses `{app_code}` directly without any prefix (e.g., `FreeBike/business-analyse/v1.0/feature.json`).
-
-**Update logic:**
-
-1. Read existing manifest at `docs/index.json` (or create empty `{ "version": "1.0", "updatedAt": "", "analyses": [] }`)
-
-2. **PROJECT MODE (`workflow.mode === "project"`):**
-   a. Add/update PROJECT-level entry:
-      - Find existing entry where `projectId == {project_id}` AND `scope == "project"`
-      - If found: update `status`, `updatedAt`
-      - If not found: append new entry with `scope: "project"`, `projectId: "{project_id}"`, `path: "business-analyse/v{version}/feature.json"`
-   b. For EACH APPLICATION:
-      - Find existing entry where `appCode == {app_code}` AND `moduleCode == null` AND `version == {version}`
-      - If found: update `status`, `updatedAt`, `featureDescription`
-      - If not found: append new entry with `moduleCode: null` and `path: "{app_code}/business-analyse/v{version}/feature.json"`
-   c. For EACH MODULE within each application:
-      - Find existing entry where `appCode == {app_code}` AND `moduleCode == {module_code}` AND `version == {version}`
-      - If found: update `status`, `updatedAt`, `featureDescription`
-      - If not found: append new entry with `path: "{app_code}/{module_code}/business-analyse/v{version}/feature.json"`
-
-3. **SINGLE-APP MODE (default):**
-   a. For the APPLICATION-level feature.json:
-      - Find existing entry where `appCode == {app_code}` AND `moduleCode == null` AND `version == {version}`
-      - If found: update `status`, `updatedAt`, `featureDescription`
-      - If not found: append new entry with `moduleCode: null` and `path: "{app_code}/business-analyse/v{version}/feature.json"`
-   b. For EACH MODULE-level feature.json:
-      - Find existing entry where `appCode == {app_code}` AND `moduleCode == {module_code}` AND `version == {version}`
-      - If found: update `status`, `updatedAt`, `featureDescription`
-      - If not found: append new entry with `moduleCode: "{module_code}"` and `path: "{app_code}/{module_code}/business-analyse/v{version}/feature.json"`
-
-4. Update root `updatedAt` to current timestamp
-5. Write manifest back to `docs/index.json`
-
-**Display confirmation:**
-```
-✓ BA manifest updated: docs/index.json
-  Entries: {total_count} ({project_count} projects, {app_count} applications, {module_count} modules)
-  Web viewer: /system/docs/ba
-```
-
-**Why a manifest?**
-- The web app needs to discover available BAs without scanning the filesystem
-- Static file serving (no backend API needed)
-- Incremental updates: each handoff adds/updates only its entries
-- Consumed by the SmartStack web app BA viewer at `/system/docs/ba`
-
----
-
-### 4. Read Source Data for HTML
-
-1. Read the HTML template from skill directory: `html/ba-interactive.html` (relative to `~/.claude/skills/business-analyse/`)
-2. **SINGLE-APP MODE:** Read the master feature.json (application level)
-3. **PROJECT MODE:** Read the project feature.json + ALL application feature.json files
-4. Read EACH module feature.json (module level, across all applications in project mode)
-
-> **Reference:** Read `references/html-data-mapping.md` for the complete FEATURE_DATA and EMBEDDED_ARTIFACTS mapping specification.
-
-**Key requirements:**
-- FEATURE_DATA must be pre-populated with ALL analysis data (empty data = BUG)
-- FEATURE_DATA.**moduleSpecs** MUST be populated for EACH module across ALL applications (use cases, business rules, entities, permissions) — empty moduleSpecs = BROKEN specification tabs
-- FEATURE_DATA.**cadrage.scope** must use HTML keys (vital/important/optional/excluded) NOT feature.json keys (mustHave/shouldHave/couldHave/outOfScope)
-- FEATURE_DATA.**metadata.analysisMode** must be set from `master.metadata.analysisMode` (always "interactive")
-- EMBEDDED_ARTIFACTS contains wireframes, E2E diagrams, dependency graph
-- Both objects are serialized as JSON with 2-space indentation
-- All data mapped from master and module feature.json files — follow `html-data-mapping.md` EXACTLY
-
-**PROJECT MODE additions to FEATURE_DATA:**
-- `FEATURE_DATA.project` — project-level metadata (name, applicationCount, applicationOrder)
-- `FEATURE_DATA.applications` — array of application summaries with their modules
-- `FEATURE_DATA.modules` — flattened list of ALL modules across ALL applications (each with `applicationCode` field)
-- `FEATURE_DATA.consolidation.crossApplicationInteractions` — from project feature.json
-- FEATURE_DATA.**metadata.workflowMode** must be `"project"` (enables multi-app tabs in HTML)
-
----
-
-### 5-7. Build FEATURE_DATA, EMBEDDED_ARTIFACTS & Replace Placeholders
-
-See [references/deploy-data-build.md](../references/deploy-data-build.md) for:
-- Complete FEATURE_DATA object structure (metadata, cadrage, modules, moduleSpecs, consolidation, handoff)
-- Complete EMBEDDED_ARTIFACTS object structure (wireframes, e2eFlows, dependencyGraph)
-- 7-step build process for FEATURE_DATA
-- 4-step artifact gathering for EMBEDDED_ARTIFACTS
-- Placeholder replacement rules (8 placeholders)
-
-**Critical reminders:**
-- FEATURE_DATA.**moduleSpecs** MUST have ONE entry per module (empty = BUG)
-- FEATURE_DATA.**cadrage.scope** converts `mustHave→vital`, `shouldHave→important`, `couldHave→optional`, `outOfScope→excluded`
-- FEATURE_DATA.**metadata.analysisMode** from `master.metadata.analysisMode`
-- Also reference `references/html-data-mapping.md` for the full mapping spec
-
----
-
-### 7-bis. Generate SVG Wireframes (Parallel Enrichment)
-
-> **After EMBEDDED_ARTIFACTS is built (sections 5-7) but BEFORE writing the HTML (section 8).**
-> This step enriches each ASCII wireframe with a professional SVG version for dual-view rendering.
-
-See [references/wireframe-svg-style-guide.md](../references/wireframe-svg-style-guide.md) for the complete SVG style specification, prompt template, and orchestration process.
-
-**Process summary:**
-
-1. **Read** `references/wireframe-svg-style-guide.md` to get the prompt template
-2. **Collect** all wireframes across ALL modules in `EMBEDDED_ARTIFACTS.wireframes` where `content` exists and `svgContent` is null
-3. **Spawn parallel Task(sonnet) agents** — ONE per wireframe, ALL in a single message
-4. **Collect and validate** results: strip markdown fences if present, verify SVG starts with `<svg` and contains `</svg>`
-5. **Inject** valid SVGs into `EMBEDDED_ARTIFACTS.wireframes[moduleCode][index].svgContent`
-6. **Display summary:**
-   ```
-   SVG wireframe generation:
-     Total wireframes: {total}
-     SVG generated: {successCount}/{total}
-     Fallback (ASCII only): {failCount}/{total}
-   ```
-
-> **CRITICAL:** This step is NEVER blocking. If all SVG generations fail, deployment
-> continues with ASCII-only wireframes. The HTML renderer checks for `svgContent` before
-> showing the SVG view. SVG is an enhancement, not a requirement.
-
----
-
-### 8. Write and Verify
-
-1. Write the populated HTML to the output directory: `docs/{app}/business-analyse/v{version}/ba-interactive.html`
-2. Display deployment confirmation:
-
-```
-[OK] Interactive HTML deployed:
-  Path: docs/{app}/business-analyse/v{version}/ba-interactive.html
-  Pre-populated with: {stakeholder_count} stakeholders, {module_count} modules,
-    {total_uc} use cases, {total_br} business rules, {total_entity} entities
-  Visual artifacts: {total_wireframes} wireframes, {e2e_flow_count} E2E diagrams
-  Open in browser to review and edit the business analysis.
-  Export JSON and re-import with: /business-analyse -x <exported-json-path>
-```
-
----
-
-## SELF-VERIFICATION (BEFORE COMPLETION)
-
-Before displaying completion, VERIFY:
-
-1. **`.ralph/prd-{module}.json`** exists for ALL modules (file size > 100 bytes each)
-2. **`.ralph/progress.txt`** exists (file size > 500 bytes)
-3. **`docs/index.json`** exists with correct entry count
-4. **ba-interactive.html** exists and file size > 100KB
-5. **FEATURE_DATA contains moduleSpecs** with entries for ALL modules
-6. **cadrage.scope uses HTML keys** (vital/important/optional/excluded)
-
-**IF any check fails → GENERATE the missing artifact before proceeding to completion.**
-
----
-
-## POST-CHECK (BLOCKING)
-
-After writing the HTML file, verify:
-
-1. **File size > 100KB** (indicates pre-populated, not empty template)
-   ```bash
-   file_size=$(stat --format=%s "{output_path}/ba-interactive.html")
-   if [ $file_size -lt 100000 ]; then
-     BLOCKING_ERROR("ba-interactive.html is only ${file_size} bytes — not pre-populated")
-   fi
-   ```
-
-2. **Grep for "moduleSpecs"** — MUST be present in generated HTML
-   ```bash
-   if ! grep -q '"moduleSpecs"' "{output_path}/ba-interactive.html"; then
-     BLOCKING_ERROR("moduleSpecs not found in HTML — FEATURE_DATA not serialized correctly")
-   fi
-   ```
-
-3. **moduleSpecs must have entries for ALL modules** (not empty object `{}`)
-   ```bash
-   # Count moduleSpecs entries — should equal module count
-   moduleSpec_count=$(grep -o '"moduleCode"' "{output_path}/ba-interactive.html" | wc -l)
-   if [ $moduleSpec_count -lt {module_count} ]; then
-     BLOCKING_ERROR("moduleSpecs has only ${moduleSpec_count} entries but need {module_count}")
-   fi
-   ```
-
-4. **cadrage.scope must use HTML keys** (vital/important/optional/excluded)
-   ```bash
-   if grep -q '"mustHave"' "{output_path}/ba-interactive.html"; then
-     BLOCKING_ERROR("cadrage.scope still has feature.json keys (mustHave) — must convert to HTML keys (vital)")
-   fi
-   ```
-
-5. **EMBEDDED_ARTIFACTS wireframes must use renamed fields** (`format`/`content`, NOT `mockupFormat`/`mockup`)
-   ```bash
-   # Check that wireframe field rename was applied (mockup → content, mockupFormat → format)
-   if grep -q '"mockupFormat"' "{output_path}/ba-interactive.html"; then
-     BLOCKING_ERROR("EMBEDDED_ARTIFACTS wireframes still use 'mockupFormat' — must rename to 'format'")
-     BLOCKING_ERROR("See references/html-data-mapping.md for correct field mapping")
-   fi
-   if grep -q '"mockup":' "{output_path}/ba-interactive.html" | grep -v '"mockupNotes"'; then
-     # Note: "mockup" in EMBEDDED_ARTIFACTS = wrong, must be "content"
-     BLOCKING_ERROR("EMBEDDED_ARTIFACTS wireframes still use 'mockup' — must rename to 'content'")
-   fi
-   ```
-
-6. **EMBEDDED_ARTIFACTS wireframes must exist for ALL modules** (not empty `{}`)
-   ```bash
-   # Each module should have at least one wireframe with "format" and "content"
-   if ! grep -q '"format"' "{output_path}/ba-interactive.html"; then
-     BLOCKING_ERROR("No wireframe 'format' field found — EMBEDDED_ARTIFACTS wireframes not populated")
-   fi
-   ```
-
-7. **WIREFRAME KEY ALIGNMENT** — wireframe keys MUST match module codes exactly
-   ```
-   FOR each module in FEATURE_DATA.modules:
-     moduleCode = module.code  (e.g., "GestionTemps")
-     IF EMBEDDED_ARTIFACTS.wireframes[moduleCode] is undefined OR empty:
-       BLOCKING_ERROR("Wireframe key mismatch: module code '${moduleCode}' not found in EMBEDDED_ARTIFACTS.wireframes")
-       BLOCKING_ERROR("Available wireframe keys: ${Object.keys(EMBEDDED_ARTIFACTS.wireframes).join(', ')}")
-       → FIX: Rename the wireframe key to match module.code exactly
-   ```
-   > **Why this matters:** The HTML renders wireframes via `EMBEDDED_ARTIFACTS.wireframes[code]`. If the key doesn't match the module code (e.g., "TimeTracking" vs "GestionTemps"), the wireframes tab shows "Aucune maquette disponible" even though data exists.
-
-8. **MODULE COMPLETENESS** — every module must have spec data AND wireframes in HTML
-   ```
-   FOR each module in master.modules[]:
-     IF FEATURE_DATA.moduleSpecs[module.code] is undefined:
-       BLOCKING_ERROR("Module '${module.code}' missing from FEATURE_DATA.moduleSpecs")
-     IF EMBEDDED_ARTIFACTS.wireframes[module.code] is undefined:
-       BLOCKING_ERROR("Module '${module.code}' missing from EMBEDDED_ARTIFACTS.wireframes")
-     IF module.featureJsonPath is null:
-       BLOCKING_ERROR("Module '${module.code}' has no per-module feature.json (featureJsonPath is null)")
-
-   expectedModules = master.modules.map(m => m.code).sort()
-   specModules = Object.keys(FEATURE_DATA.moduleSpecs).sort()
-   wireframeModules = Object.keys(EMBEDDED_ARTIFACTS.wireframes).sort()
-
-   IF expectedModules.join(',') !== specModules.join(','):
-     BLOCKING_ERROR("Module mismatch: expected [${expectedModules}] but moduleSpecs has [${specModules}]")
-   IF expectedModules.join(',') !== wireframeModules.join(','):
-     BLOCKING_ERROR("Module mismatch: expected [${expectedModules}] but wireframes has [${wireframeModules}]")
-   ```
-
-9. **MODULE DATA COMPLETENESS** — every module in FEATURE_DATA.modules[] must carry ALL fields from feature.json
-   ```
-   FOR each htmlModule in FEATURE_DATA.modules[]:
-     masterModule = master.modules.find(m => m.code === htmlModule.code)
-
-     // name MUST be display name, NOT code
-     IF htmlModule.name === htmlModule.code AND masterModule.name !== masterModule.code:
-       BLOCKING_ERROR("Module '${htmlModule.code}' has name === code. Expected: '${masterModule.name}'")
-
-     // anticipatedSections MUST be present (drives Structure tab + navigation tree)
-     IF masterModule.anticipatedSections?.length > 0 AND (!htmlModule.anticipatedSections OR htmlModule.anticipatedSections.length === 0):
-       BLOCKING_ERROR("Module '${htmlModule.code}' lost anticipatedSections: master has ${masterModule.anticipatedSections.length}, HTML has 0")
-       → FIX: Copy anticipatedSections from master.modules[].anticipatedSections
-
-     // dependencies/dependents MUST be present (drives dependency graph)
-     IF masterModule.dependencies?.length > 0 AND (!htmlModule.dependencies OR htmlModule.dependencies.length === 0):
-       BLOCKING_ERROR("Module '${htmlModule.code}' lost dependencies array")
-   ```
-
-   Display verification table:
-   ```
-   POST-CHECK: Module data completeness
-   | Module         | name    | sections | deps | wireframes | specs | Match |
-   |----------------|---------|----------|------|------------|-------|-------|
-   | {code}         | {name}  | {n}      | {n}  | {n}        | OK    | OK/FAIL |
-   ```
-
-**IF ANY CHECK FAILS → DO NOT PROCEED. Fix the data mapping and regenerate.**
-
----
-
-## COMPLETION
-
-> **IMPORTANT:** La BA ne lance JAMAIS /ralph-loop automatiquement.
-> Il peut y avoir une phase de revue/modification de la documentation entre les deux étapes.
-> L'utilisateur lance `/ralph-loop` manuellement quand il est prêt.
-
-Display completion using the template from `templates/tpl-launch-displays.md`:
-
-```
-═══════════════════════════════════════════════════════════════
-  [OK] BUSINESS ANALYSE TERMINEE - {application_name OR project_name}
-═══════════════════════════════════════════════════════════════
-
-{IF workflow.mode === "project":}
-Project:   {project_name}
-Apps:      {app_count} ({app_names})
-{ENDIF}
-Modules:   {count} ({names})
-Strategy:  {strategy}
-Files:     {total files across all modules}
-Tasks:     {total tasks} ({core_count} CORE + {biz_count} business + {dev_count} development)
-Complexity: {complexity}
-Effort:    {total_days} days ({total_hours} hours)
-
-[DIR] Artefacts générés:
-  ✓ feature.json ({IF project: "project + " ENDIF}master + per-module) - spécification complète
-  ✓ .ralph/prd-{module}.json - task breakdown per module
-  {IF workflow.mode === "project":}
-  ✓ .ralph/prd-CrossApplication.json - cross-application integration tests
-  {ENDIF}
-  ✓ .ralph/progress.txt - tracker de progression
-  ✓ ba-interactive.html - document de revue interactif
-
-[TARGET] Prochaines étapes:
-  1. Ouvrir ba-interactive.html dans le navigateur
-  2. Partager avec les stakeholders pour validation
-  3. Si retours --> relancer /business-analyse pour une nouvelle itération
-  4. Validation qualité ULTRATHINK (automatique):
-
-     → Lancement automatique de step-05c-ralph-readiness.md...
-
-  5. Une fois validé, lancer le développement:
-
-     /ralph-loop
-
-     Note: /ralph-loop -r sert UNIQUEMENT à reprendre une boucle interrompue.
-           Après un /business-analyse, toujours utiliser /ralph-loop (sans -r).
-
-═══════════════════════════════════════════════════════════════
-```
-
----
-
-## NEXT STEP
-
-**MANDATORY:** Load `steps/step-05c-ralph-readiness.md` — ULTRATHINK quality review & readiness gate.
-
-> **This step is NON-NEGOTIABLE.** The BA skill MUST run step-05c before completing.
-> DO NOT skip this step. DO NOT display a "done" message without running step-05c first.
-
-This validation ensures:
-- ULTRATHINK deep quality review of the entire BA output
-- All module handoffs are complete and content is coherent
-- PRD files are structurally valid
-- Cross-module references are resolvable
-- Specification quality meets production standards
-- No blocking issues before development
-
----
-
-## MODE SUPPORT & TROUBLESHOOTING
-
-See [references/deploy-modes.md](../references/deploy-modes.md) for:
-- **Standard Mode** — full handoff (all 7 categories, complete mapping)
-- **Micro Mode** — simplified CRUD (3 core SeedData, lightweight)
-- **Delta Mode** — refactoring focus (affected modules only)
-- **Troubleshooting** — 12 common issues with resolutions
+---
+name: step-05b-deploy
+description: Generate prd.json, deploy artifacts, interactive HTML
+model: sonnet
+next_step: steps/step-05c-ralph-readiness.md
+---
+
+> **Context files:** _shared.md
+
+# Step 5b: Deploy Artifacts & Interactive HTML
+
+## MANDATORY EXECUTION RULES
+
+- **ALWAYS** verify all module handoffs are complete before generating artifacts
+- **ALWAYS** derive prd.json from feature.json (NEVER independently)
+- **NEVER** invent entities/FRs/BRs not in feature.json
+- **ALWAYS** update BA manifest at docs/index.json
+- **ALWAYS** deploy ba-interactive.html PRE-POPULATED with all data
+- **ALWAYS** follow references/html-data-mapping.md EXACTLY for FEATURE_DATA and EMBEDDED_ARTIFACTS mapping
+- **FEATURE_DATA.moduleSpecs MUST be populated for EACH module** — empty moduleSpecs = BROKEN specification tabs
+- **FEATURE_DATA.cadrage.scope must use HTML keys** (vital/important/optional/excluded) NOT feature.json keys (mustHave/shouldHave/couldHave/outOfScope)
+- **FEATURE_DATA.metadata.analysisMode must be set** from master.metadata.analysisMode (always "interactive")
+- **NEVER leave FEATURE_DATA or EMBEDDED_ARTIFACTS empty or partial**
+
+## YOUR TASK
+
+Generate deployment artifacts (prd.json, progress.txt, BA manifest) and deploy the interactive HTML document pre-populated with all analysis data.
+
+---
+
+## EXECUTION SEQUENCE
+
+### 0. Pre-flight Verification
+
+Before generating ANY artifact, verify step-05a completed successfully:
+
+```
+FOR each module in modules[]:
+  Read module feature.json
+  IF module.handoff === {} OR module.status !== "handed-off":
+    → BLOCKING ERROR: Module {module.code} has no handoff data
+    → Return to step-05a-handoff.md
+```
+
+IF all modules have valid handoff → proceed.
+
+---
+
+### 1. Generate prd.json (PROGRAMMATIC)
+
+> **RULE:** prd.json is extracted by CLI code, **NEVER** generated by LLM.
+> The `ss derive-prd` command performs a deterministic data transformation from feature.json.
+
+**For each module:**
+
+```
+Execute: ss derive-prd --feature {moduleFeaturePath} --output .ralph/prd-{moduleCode}.json
+```
+
+**For consolidated view (multi-module, optional):**
+
+```
+Execute: ss derive-prd --application {masterFeaturePath}
+→ Generates .ralph/prd-{moduleCode}.json for each module
+```
+
+**Verification:** After execution, read the generated prd.json and display summary:
+
+```
+prd.json generated for module {moduleCode}:
+  - Use cases: {count}
+  - Functional requirements: {count}
+  - Business rules: {count}
+  - API endpoints: {count}
+  - Sections: {count}
+  - Files to create: {count}
+  - BR-to-code mappings: {count}
+```
+
+**Key guarantees:**
+- Source MUST reference feature.json path (traceability)
+- All data is EXACT COPY from feature.json (no transformation, no invention)
+- prd.json $version: "3.0.0" (unified format with pre-computed tasks)
+- source.type: "ba-handoff-programmatic"
+
+**POST-CHECK (BLOCKING — DO NOT SKIP):**
+
+After `ss derive-prd` execution, read the generated prd-{moduleCode}.json and verify structural integrity:
+
+```javascript
+const prd = readJSON(`.ralph/prd-${moduleCode}.json`);
+const featureHandoff = moduleFeature.handoff.filesToCreate;
+
+// 1. Verify structure: filesToCreate MUST be under implementation (NOT root level)
+if (prd.filesToCreate && !prd.implementation?.filesToCreate) {
+  BLOCKING_ERROR("prd.json has filesToCreate at ROOT level — this is NOT from ss derive-prd");
+  BLOCKING_ERROR("Re-run: ss derive-prd --feature {path} --output .ralph/prd-{moduleCode}.json");
+  STOP;
+}
+
+// 2. Verify ALL 7 categories present and match feature.json
+const categories = ['domain', 'application', 'infrastructure', 'api', 'frontend', 'seedData', 'tests'];
+for (const cat of categories) {
+  const prdCount = prd.implementation.filesToCreate[cat]?.length ?? 0;
+  const featureCount = featureHandoff[cat]?.length ?? 0;
+  if (prdCount !== featureCount) {
+    BLOCKING_ERROR(`${cat}: prd has ${prdCount} files but feature.json has ${featureCount}`);
+  }
+}
+```
+
+Display verification table:
+
+```
+POST-CHECK: prd-{moduleCode}.json integrity
+| Category       | feature.json | prd.json | Match |
+|----------------|-------------|----------|-------|
+| domain         | {n}         | {n}      | OK/FAIL |
+| application    | {n}         | {n}      | OK/FAIL |
+| infrastructure | {n}         | {n}      | OK/FAIL |
+| api            | {n}         | {n}      | OK/FAIL |
+| frontend       | {n}         | {n}      | OK/FAIL |
+| seedData       | {n}         | {n}      | OK/FAIL |
+| tests          | {n}         | {n}      | OK/FAIL |
+```
+
+IF ANY category shows FAIL → **STOP AND RE-RUN `ss derive-prd`**. DO NOT proceed with incomplete PRD.
+
+---
+
+### 2. Initialize Progress Tracker
+
+> **Template:** Read `templates/tpl-progress.md` for the complete progress tracker template structure.
+> Populate the template with module-specific data from feature.json handoff sections.
+
+**Progress Tracker Rules:**
+- One section per module, in topological order (dependencies first)
+- CORE SeedData ALWAYS 5 entries (mandatory)
+- Business SeedData varies by module
+- Hierarchical task structure (module -> layer -> tasks)
+- Each task is independent, assignable checkbox
+- Effort estimate per module (simple/medium/complex)
+- Summary with totals across all modules
+- Cross-module tasks only if multi-module
+
+---
+
+### 3. Update BA Manifest (MANDATORY)
+
+> **The BA manifest enables the SmartStack web app to discover and display all available business analyses.**
+> It is a JSON index file at `docs/index.json` that lists all feature.json files.
+
+**Path:** `docs/index.json` (project root relative)
+
+**Schema:**
+```json
+{
+  "version": "1.0",
+  "updatedAt": "{ISO timestamp}",
+  "analyses": [
+    {
+      "appCode": "{app_code}",
+      "appName": "{application_name}",
+      "moduleCode": null | "{module_code}",
+      "moduleName": "{module_name}",
+      "version": "{version}",
+      "status": "handed-off",
+      "featureDescription": "{feature_description}",
+      "path": "{app_code}/business-analyse/v{version}/feature.json",
+      "updatedAt": "{ISO timestamp}"
+    }
+  ]
+}
+```
+
+> Note: The "path" field uses `{app_code}` directly without any prefix (e.g., `FreeBike/business-analyse/v1.0/feature.json`).
+
+**Update logic:**
+
+1. Read existing manifest at `docs/index.json` (or create empty `{ "version": "1.0", "updatedAt": "", "analyses": [] }`)
+
+2. **PROJECT MODE (`workflow.mode === "project"`):**
+   a. Add/update PROJECT-level entry:
+      - Find existing entry where `projectId == {project_id}` AND `scope == "project"`
+      - If found: update `status`, `updatedAt`
+      - If not found: append new entry with `scope: "project"`, `projectId: "{project_id}"`, `path: "business-analyse/v{version}/feature.json"`
+   b. For EACH APPLICATION:
+      - Find existing entry where `appCode == {app_code}` AND `moduleCode == null` AND `version == {version}`
+      - If found: update `status`, `updatedAt`, `featureDescription`
+      - If not found: append new entry with `moduleCode: null` and `path: "{app_code}/business-analyse/v{version}/feature.json"`
+   c. For EACH MODULE within each application:
+      - Find existing entry where `appCode == {app_code}` AND `moduleCode == {module_code}` AND `version == {version}`
+      - If found: update `status`, `updatedAt`, `featureDescription`
+      - If not found: append new entry with `path: "{app_code}/{module_code}/business-analyse/v{version}/feature.json"`
+
+3. **SINGLE-APP MODE (default):**
+   a. For the APPLICATION-level feature.json:
+      - Find existing entry where `appCode == {app_code}` AND `moduleCode == null` AND `version == {version}`
+      - If found: update `status`, `updatedAt`, `featureDescription`
+      - If not found: append new entry with `moduleCode: null` and `path: "{app_code}/business-analyse/v{version}/feature.json"`
+   b. For EACH MODULE-level feature.json:
+      - Find existing entry where `appCode == {app_code}` AND `moduleCode == {module_code}` AND `version == {version}`
+      - If found: update `status`, `updatedAt`, `featureDescription`
+      - If not found: append new entry with `moduleCode: "{module_code}"` and `path: "{app_code}/{module_code}/business-analyse/v{version}/feature.json"`
+
+4. Update root `updatedAt` to current timestamp
+5. Write manifest back to `docs/index.json`
+
+**Display confirmation:**
+```
+✓ BA manifest updated: docs/index.json
+  Entries: {total_count} ({project_count} projects, {app_count} applications, {module_count} modules)
+  Web viewer: /system/docs/ba
+```
+
+**Why a manifest?**
+- The web app needs to discover available BAs without scanning the filesystem
+- Static file serving (no backend API needed)
+- Incremental updates: each handoff adds/updates only its entries
+- Consumed by the SmartStack web app BA viewer at `/system/docs/ba`
+
+---
+
+### 4. Read Source Data for HTML
+
+1. Read the HTML template from skill directory: `html/ba-interactive.html` (relative to `~/.claude/skills/business-analyse/`)
+2. **SINGLE-APP MODE:** Read the master feature.json (application level)
+3. **PROJECT MODE:** Read the project feature.json + ALL application feature.json files
+4. Read EACH module feature.json (module level, across all applications in project mode)
+
+> **Reference:** Read `references/html-data-mapping.md` for the complete FEATURE_DATA and EMBEDDED_ARTIFACTS mapping specification.
+
+**Key requirements:**
+- FEATURE_DATA must be pre-populated with ALL analysis data (empty data = BUG)
+- FEATURE_DATA.**moduleSpecs** MUST be populated for EACH module across ALL applications (use cases, business rules, entities, permissions) — empty moduleSpecs = BROKEN specification tabs
+- FEATURE_DATA.**cadrage.scope** must use HTML keys (vital/important/optional/excluded) NOT feature.json keys (mustHave/shouldHave/couldHave/outOfScope)
+- FEATURE_DATA.**metadata.analysisMode** must be set from `master.metadata.analysisMode` (always "interactive")
+- EMBEDDED_ARTIFACTS contains wireframes, E2E diagrams, dependency graph
+- Both objects are serialized as JSON with 2-space indentation
+- All data mapped from master and module feature.json files — follow `html-data-mapping.md` EXACTLY
+
+**PROJECT MODE additions to FEATURE_DATA:**
+- `FEATURE_DATA.project` — project-level metadata (name, applicationCount, applicationOrder)
+- `FEATURE_DATA.applications` — array of application summaries with their modules
+- `FEATURE_DATA.modules` — flattened list of ALL modules across ALL applications (each with `applicationCode` field)
+- `FEATURE_DATA.consolidation.crossApplicationInteractions` — from project feature.json
+- FEATURE_DATA.**metadata.workflowMode** must be `"project"` (enables multi-app tabs in HTML)
+
+---
+
+### 5-7. Build FEATURE_DATA, EMBEDDED_ARTIFACTS & Replace Placeholders
+
+See [references/deploy-data-build.md](../references/deploy-data-build.md) for:
+- Complete FEATURE_DATA object structure (metadata, cadrage, modules, moduleSpecs, consolidation, handoff)
+- Complete EMBEDDED_ARTIFACTS object structure (wireframes, e2eFlows, dependencyGraph)
+- 7-step build process for FEATURE_DATA
+- 4-step artifact gathering for EMBEDDED_ARTIFACTS
+- Placeholder replacement rules (8 placeholders)
+
+**Critical reminders:**
+- FEATURE_DATA.**moduleSpecs** MUST have ONE entry per module (empty = BUG)
+- FEATURE_DATA.**cadrage.scope** converts `mustHave→vital`, `shouldHave→important`, `couldHave→optional`, `outOfScope→excluded`
+- FEATURE_DATA.**metadata.analysisMode** from `master.metadata.analysisMode`
+- Also reference `references/html-data-mapping.md` for the full mapping spec
+
+---
+
+### 7-bis. Generate SVG Wireframes (Parallel Enrichment)
+
+> **After EMBEDDED_ARTIFACTS is built (sections 5-7) but BEFORE writing the HTML (section 8).**
+> This step enriches each ASCII wireframe with a professional SVG version for dual-view rendering.
+
+See [references/wireframe-svg-style-guide.md](../references/wireframe-svg-style-guide.md) for the complete SVG style specification, prompt template, and orchestration process.
+
+**Process summary:**
+
+1. **Read** `references/wireframe-svg-style-guide.md` to get the prompt template
+2. **Collect** all wireframes across ALL modules in `EMBEDDED_ARTIFACTS.wireframes` where `content` exists and `svgContent` is null
+3. **Spawn parallel Task(sonnet) agents** — ONE per wireframe, ALL in a single message
+4. **Collect and validate** results: strip markdown fences if present, verify SVG starts with `<svg` and contains `</svg>`
+5. **Inject** valid SVGs into `EMBEDDED_ARTIFACTS.wireframes[moduleCode][index].svgContent`
+6. **Display summary:**
+   ```
+   SVG wireframe generation:
+     Total wireframes: {total}
+     SVG generated: {successCount}/{total}
+     Fallback (ASCII only): {failCount}/{total}
+   ```
+
+> **CRITICAL:** This step is NEVER blocking. If all SVG generations fail, deployment
+> continues with ASCII-only wireframes. The HTML renderer checks for `svgContent` before
+> showing the SVG view. SVG is an enhancement, not a requirement.
+
+---
+
+### 8. Write and Verify
+
+1. Write the populated HTML to the output directory: `docs/{app}/business-analyse/v{version}/ba-interactive.html`
+2. Display deployment confirmation:
+
+```
+[OK] Interactive HTML deployed:
+  Path: docs/{app}/business-analyse/v{version}/ba-interactive.html
+  Pre-populated with: {stakeholder_count} stakeholders, {module_count} modules,
+    {total_uc} use cases, {total_br} business rules, {total_entity} entities
+  Visual artifacts: {total_wireframes} wireframes, {e2e_flow_count} E2E diagrams
+  Open in browser to review and edit the business analysis.
+  Export JSON and re-import with: /business-analyse -x <exported-json-path>
+```
+
+---
+
+## SELF-VERIFICATION (BEFORE COMPLETION)
+
+Before displaying completion, VERIFY:
+
+1. **`.ralph/prd-{module}.json`** exists for ALL modules (file size > 100 bytes each)
+2. **`.ralph/progress.txt`** exists (file size > 500 bytes)
+3. **`docs/index.json`** exists with correct entry count
+4. **ba-interactive.html** exists and file size > 100KB
+5. **FEATURE_DATA contains moduleSpecs** with entries for ALL modules
+6. **cadrage.scope uses HTML keys** (vital/important/optional/excluded)
+
+**IF any check fails → GENERATE the missing artifact before proceeding to completion.**
+
+---
+
+## POST-CHECK (BLOCKING)
+
+After writing the HTML file, verify:
+
+1. **File size > 100KB** (indicates pre-populated, not empty template)
+   ```bash
+   file_size=$(stat --format=%s "{output_path}/ba-interactive.html")
+   if [ $file_size -lt 100000 ]; then
+     BLOCKING_ERROR("ba-interactive.html is only ${file_size} bytes — not pre-populated")
+   fi
+   ```
+
+2. **Grep for "moduleSpecs"** — MUST be present in generated HTML
+   ```bash
+   if ! grep -q '"moduleSpecs"' "{output_path}/ba-interactive.html"; then
+     BLOCKING_ERROR("moduleSpecs not found in HTML — FEATURE_DATA not serialized correctly")
+   fi
+   ```
+
+3. **moduleSpecs must have entries for ALL modules** (not empty object `{}`)
+   ```bash
+   # Count moduleSpecs entries — should equal module count
+   moduleSpec_count=$(grep -o '"moduleCode"' "{output_path}/ba-interactive.html" | wc -l)
+   if [ $moduleSpec_count -lt {module_count} ]; then
+     BLOCKING_ERROR("moduleSpecs has only ${moduleSpec_count} entries but need {module_count}")
+   fi
+   ```
+
+4. **cadrage.scope must use HTML keys** (vital/important/optional/excluded)
+   ```bash
+   if grep -q '"mustHave"' "{output_path}/ba-interactive.html"; then
+     BLOCKING_ERROR("cadrage.scope still has feature.json keys (mustHave) — must convert to HTML keys (vital)")
+   fi
+   ```
+
+5. **EMBEDDED_ARTIFACTS wireframes must use renamed fields** (`format`/`content`, NOT `mockupFormat`/`mockup`)
+   ```bash
+   # Check that wireframe field rename was applied (mockup → content, mockupFormat → format)
+   if grep -q '"mockupFormat"' "{output_path}/ba-interactive.html"; then
+     BLOCKING_ERROR("EMBEDDED_ARTIFACTS wireframes still use 'mockupFormat' — must rename to 'format'")
+     BLOCKING_ERROR("See references/html-data-mapping.md for correct field mapping")
+   fi
+   if grep -q '"mockup":' "{output_path}/ba-interactive.html" | grep -v '"mockupNotes"'; then
+     # Note: "mockup" in EMBEDDED_ARTIFACTS = wrong, must be "content"
+     BLOCKING_ERROR("EMBEDDED_ARTIFACTS wireframes still use 'mockup' — must rename to 'content'")
+   fi
+   ```
+
+6. **EMBEDDED_ARTIFACTS wireframes must exist for ALL modules** (not empty `{}`)
+   ```bash
+   # Each module should have at least one wireframe with "format" and "content"
+   if ! grep -q '"format"' "{output_path}/ba-interactive.html"; then
+     BLOCKING_ERROR("No wireframe 'format' field found — EMBEDDED_ARTIFACTS wireframes not populated")
+   fi
+   ```
+
+7. **WIREFRAME KEY ALIGNMENT** — wireframe keys MUST match module codes exactly
+   ```
+   FOR each module in FEATURE_DATA.modules:
+     moduleCode = module.code  (e.g., "GestionTemps")
+     IF EMBEDDED_ARTIFACTS.wireframes[moduleCode] is undefined OR empty:
+       BLOCKING_ERROR("Wireframe key mismatch: module code '${moduleCode}' not found in EMBEDDED_ARTIFACTS.wireframes")
+       BLOCKING_ERROR("Available wireframe keys: ${Object.keys(EMBEDDED_ARTIFACTS.wireframes).join(', ')}")
+       → FIX: Rename the wireframe key to match module.code exactly
+   ```
+   > **Why this matters:** The HTML renders wireframes via `EMBEDDED_ARTIFACTS.wireframes[code]`. If the key doesn't match the module code (e.g., "TimeTracking" vs "GestionTemps"), the wireframes tab shows "Aucune maquette disponible" even though data exists.
+
+8. **MODULE COMPLETENESS** — every module must have spec data AND wireframes in HTML
+   ```
+   FOR each module in master.modules[]:
+     IF FEATURE_DATA.moduleSpecs[module.code] is undefined:
+       BLOCKING_ERROR("Module '${module.code}' missing from FEATURE_DATA.moduleSpecs")
+     IF EMBEDDED_ARTIFACTS.wireframes[module.code] is undefined:
+       BLOCKING_ERROR("Module '${module.code}' missing from EMBEDDED_ARTIFACTS.wireframes")
+     IF module.featureJsonPath is null:
+       BLOCKING_ERROR("Module '${module.code}' has no per-module feature.json (featureJsonPath is null)")
+
+   expectedModules = master.modules.map(m => m.code).sort()
+   specModules = Object.keys(FEATURE_DATA.moduleSpecs).sort()
+   wireframeModules = Object.keys(EMBEDDED_ARTIFACTS.wireframes).sort()
+
+   IF expectedModules.join(',') !== specModules.join(','):
+     BLOCKING_ERROR("Module mismatch: expected [${expectedModules}] but moduleSpecs has [${specModules}]")
+   IF expectedModules.join(',') !== wireframeModules.join(','):
+     BLOCKING_ERROR("Module mismatch: expected [${expectedModules}] but wireframes has [${wireframeModules}]")
+   ```
+
+9. **MODULE DATA COMPLETENESS** — every module in FEATURE_DATA.modules[] must carry ALL fields from feature.json
+   ```
+   FOR each htmlModule in FEATURE_DATA.modules[]:
+     masterModule = master.modules.find(m => m.code === htmlModule.code)
+
+     // name MUST be display name, NOT code
+     IF htmlModule.name === htmlModule.code AND masterModule.name !== masterModule.code:
+       BLOCKING_ERROR("Module '${htmlModule.code}' has name === code. Expected: '${masterModule.name}'")
+
+     // anticipatedSections MUST be present (drives Structure tab + navigation tree)
+     IF masterModule.anticipatedSections?.length > 0 AND (!htmlModule.anticipatedSections OR htmlModule.anticipatedSections.length === 0):
+       BLOCKING_ERROR("Module '${htmlModule.code}' lost anticipatedSections: master has ${masterModule.anticipatedSections.length}, HTML has 0")
+       → FIX: Copy anticipatedSections from master.modules[].anticipatedSections
+
+     // dependencies/dependents MUST be present (drives dependency graph)
+     IF masterModule.dependencies?.length > 0 AND (!htmlModule.dependencies OR htmlModule.dependencies.length === 0):
+       BLOCKING_ERROR("Module '${htmlModule.code}' lost dependencies array")
+   ```
+
+   Display verification table:
+   ```
+   POST-CHECK: Module data completeness
+   | Module         | name    | sections | deps | wireframes | specs | Match |
+   |----------------|---------|----------|------|------------|-------|-------|
+   | {code}         | {name}  | {n}      | {n}  | {n}        | OK    | OK/FAIL |
+   ```
+
+**IF ANY CHECK FAILS → DO NOT PROCEED. Fix the data mapping and regenerate.**
+
+---
+
+## COMPLETION
+
+> **IMPORTANT:** La BA ne lance JAMAIS /ralph-loop automatiquement.
+> Il peut y avoir une phase de revue/modification de la documentation entre les deux étapes.
+> L'utilisateur lance `/ralph-loop` manuellement quand il est prêt.
+
+Display completion using the template from `templates/tpl-launch-displays.md`:
+
+```
+═══════════════════════════════════════════════════════════════
+  [OK] BUSINESS ANALYSE TERMINEE - {application_name OR project_name}
+═══════════════════════════════════════════════════════════════
+
+{IF workflow.mode === "project":}
+Project:   {project_name}
+Apps:      {app_count} ({app_names})
+{ENDIF}
+Modules:   {count} ({names})
+Strategy:  {strategy}
+Files:     {total files across all modules}
+Tasks:     {total tasks} ({core_count} CORE + {biz_count} business + {dev_count} development)
+Complexity: {complexity}
+Effort:    {total_days} days ({total_hours} hours)
+
+[DIR] Artefacts générés:
+  ✓ feature.json ({IF project: "project + " ENDIF}master + per-module) - spécification complète
+  ✓ .ralph/prd-{module}.json - task breakdown per module
+  {IF workflow.mode === "project":}
+  ✓ .ralph/prd-CrossApplication.json - cross-application integration tests
+  {ENDIF}
+  ✓ .ralph/progress.txt - tracker de progression
+  ✓ ba-interactive.html - document de revue interactif
+
+[TARGET] Prochaines étapes:
+  1. Ouvrir ba-interactive.html dans le navigateur
+  2. Partager avec les stakeholders pour validation
+  3. Si retours --> relancer /business-analyse pour une nouvelle itération
+  4. Validation qualité ULTRATHINK (automatique):
+
+     → Lancement automatique de step-05c-ralph-readiness.md...
+
+  5. Une fois validé, lancer le développement:
+
+     /ralph-loop
+
+     Note: /ralph-loop -r sert UNIQUEMENT à reprendre une boucle interrompue.
+           Après un /business-analyse, toujours utiliser /ralph-loop (sans -r).
+
+═══════════════════════════════════════════════════════════════
+```
+
+---
+
+## NEXT STEP
+
+**MANDATORY:** Load `steps/step-05c-ralph-readiness.md` — ULTRATHINK quality review & readiness gate.
+
+> **This step is NON-NEGOTIABLE.** The BA skill MUST run step-05c before completing.
+> DO NOT skip this step. DO NOT display a "done" message without running step-05c first.
+
+This validation ensures:
+- ULTRATHINK deep quality review of the entire BA output
+- All module handoffs are complete and content is coherent
+- PRD files are structurally valid
+- Cross-module references are resolvable
+- Specification quality meets production standards
+- No blocking issues before development
+
+---
+
+## MODE SUPPORT & TROUBLESHOOTING
+
+See [references/deploy-modes.md](../references/deploy-modes.md) for:
+- **Standard Mode** — full handoff (all 7 categories, complete mapping)
+- **Micro Mode** — simplified CRUD (3 core SeedData, lightweight)
+- **Delta Mode** — refactoring focus (affected modules only)
+- **Troubleshooting** — 12 common issues with resolutions