@atlashub/smartstack-cli 4.81.0 → 5.0.0

package/templates/skills/apex/steps/step-07-tests.md

@@ -1,271 +0,0 @@
----
-name: step-07-tests
-description: Final Test Sweep — security tests, coverage analysis, remaining failures
-model: opus
-prev_step: steps/step-06-resolve.md
-next_step: steps/step-08-run-tests.md
----
-
-# Step 7: Final Test Sweep
-
-> **MANDATORY STEP — This is NOT optional.** APEX requires a final test sweep for completion.
-> Backend and frontend tests were already scaffolded and run inline (Layer 2 and Layer 3).
-> This step focuses on: security tests, coverage gaps, cross-layer scenarios, and remaining failures.
-
-**Goal:** Ensure comprehensive test coverage. Target: >= 80% coverage, 100% pass rate. Acceptable fallback: skip failing tests with TODO comment after 3 fix attempts — escalate to user rather than blocking indefinitely.
-
----
-
-## Task Progress
-
-### Progress Tracker Update
-```
-TaskUpdate(taskId: progress_tracker_id,
-  description: "Module: {module_code}. Current: step-07 (Final Test Sweep)",
-  activeForm: "Running test sweep")
-```
-
----
-
-## 0. Ensure Test Infrastructure Exists
-
-> **Note:** These infrastructure checks are idempotent safety nets. If inline tests (step-03 Layer 2/3) ran successfully, the infrastructure already exists. These checks guard against edge cases where inline test scaffolding was skipped or partially completed.
-
-### 0a. Frontend Test Tooling (if frontend was generated)
-
-> **Before scaffolding additional frontend tests, the test infrastructure must be in place.**
-
-```bash
-# Check if Vitest is installed
-WEB_DIR=$(find . -name "package.json" -not -path "*/node_modules/*" -exec dirname {} \; 2>/dev/null | head -1)
-if [ -n "$WEB_DIR" ]; then
-  cd "$WEB_DIR"
-
-  # Install test dependencies if missing
-  if ! grep -q '"vitest"' package.json 2>/dev/null; then
-    npm install -D vitest @testing-library/react @testing-library/jest-dom @testing-library/user-event jsdom @types/testing-library__jest-dom
-  fi
-
-  # Create vitest.config.ts if missing
-  if [ ! -f "vitest.config.ts" ]; then
-    cat > vitest.config.ts << 'VITEST_EOF'
-import { defineConfig } from 'vitest/config';
-import react from '@vitejs/plugin-react';
-import path from 'path';
-
-export default defineConfig({
-  plugins: [react()],
-  test: {
-    globals: true,
-    environment: 'jsdom',
-    setupFiles: ['./src/test/setup.ts'],
-    css: true,
-    include: ['src/**/*.test.{ts,tsx}'],
-  },
-  resolve: {
-    alias: {
-      '@': path.resolve(__dirname, './src'),
-    },
-  },
-});
-VITEST_EOF
-  fi
-
-  # Create test setup file if missing
-  mkdir -p src/test
-  if [ ! -f "src/test/setup.ts" ]; then
-    cat > src/test/setup.ts << 'SETUP_EOF'
-import '@testing-library/jest-dom';
-SETUP_EOF
-  fi
-
-  # Add test script to package.json if missing
-  if ! grep -q '"test"' package.json 2>/dev/null; then
-    # Use npm pkg set for safe JSON manipulation
-    npm pkg set scripts.test="vitest run"
-    npm pkg set scripts.test:watch="vitest"
-    npm pkg set scripts.test:coverage="vitest run --coverage"
-  fi
-
-  cd -
-fi
-```
-
-### 0b. Backend Test Projects
-
-```bash
-# Check for existing unit test project
-UNIT_PROJECT=$(find tests/ -name "*.Tests.Unit.csproj" 2>/dev/null | head -1)
-
-if [ -z "$UNIT_PROJECT" ]; then
-  PROJECT_NAME=$(basename *.sln .sln)
-  dotnet new xunit -n "${PROJECT_NAME}.Tests.Unit" -o "tests/${PROJECT_NAME}.Tests.Unit"
-  dotnet add "tests/${PROJECT_NAME}.Tests.Unit" package Moq
-  dotnet add "tests/${PROJECT_NAME}.Tests.Unit" package FluentAssertions
-  for proj in src/*/*.csproj; do
-    dotnet add "tests/${PROJECT_NAME}.Tests.Unit" reference "$proj"
-  done
-  dotnet sln add "tests/${PROJECT_NAME}.Tests.Unit/${PROJECT_NAME}.Tests.Unit.csproj"
-fi
-```
-
-### 0c. Integration Test Project (for DB validation)
-
-```bash
-# Check for existing integration test project
-INT_PROJECT=$(find tests/ -name "*.Tests.Integration.csproj" 2>/dev/null | head -1)
-
-if [ -z "$INT_PROJECT" ]; then
-  PROJECT_NAME=$(basename *.sln .sln)
-  dotnet new xunit -n "${PROJECT_NAME}.Tests.Integration" -o "tests/${PROJECT_NAME}.Tests.Integration"
-  dotnet add "tests/${PROJECT_NAME}.Tests.Integration" package Moq
-  dotnet add "tests/${PROJECT_NAME}.Tests.Integration" package FluentAssertions
-  dotnet add "tests/${PROJECT_NAME}.Tests.Integration" package Respawn
-  dotnet add "tests/${PROJECT_NAME}.Tests.Integration" package Microsoft.EntityFrameworkCore.SqlServer
-  for proj in src/*/*.csproj; do
-    dotnet add "tests/${PROJECT_NAME}.Tests.Integration" reference "$proj"
-  done
-  dotnet sln add "tests/${PROJECT_NAME}.Tests.Integration/${PROJECT_NAME}.Tests.Integration.csproj"
-fi
-```
-
-> **Note:** Integration tests use `DatabaseFixture` + `Respawn` with SQL Server LocalDB.
-> See step-04 section 5c for details.
-
----
-
-## 1. Conditional Skip — Inline Tests Already Scaffolded
-
-> **Backend and frontend tests were already scaffolded inline in step-03 (Layer 2 and Layer 3).**
-> Do NOT re-scaffold tests that already exist. Only scaffold NEW test types here.
-
-```
-Check existing tests:
-  → If domain tests exist (from Layer 2 inline): SKIP domain scaffold
-  → If application tests exist (from Layer 2 inline): SKIP application scaffold
-  → If API integration tests exist (from Layer 2 inline): SKIP API scaffold
-  → If frontend form tests exist (from Layer 3 inline): SKIP frontend form scaffold
-```
-
----
-
-## 2. Security Tests (PRIMARY focus of this step)
-
-```
-Call: mcp__smartstack__scaffold_tests
-  module: "{module_code}"
-  test_type: "security"
-
-Tests: RequirePermission enforcement, multi-tenant isolation, tenant boundary violations
-```
-
-> **Security tests are the main value-add of step-07.** Inline tests cover CRUD functionality;
-> security tests verify authorization and isolation — these require cross-layer awareness.
-
----
-
-## 3. Scaffold Missing Tests (if any layers skipped inline)
-
-For each layer that was NOT tested inline:
-
-### Domain Tests (if not already scaffolded)
-```
-Call: mcp__smartstack__scaffold_tests
-  target_layer: "domain"
-  module: "{module_code}"
-  test_type: "unit"
-
-Tests: entity creation, validation, domain events, value objects
-```
-
-### Application Tests (if not already scaffolded)
-```
-Call: mcp__smartstack__scaffold_tests
-  target_layer: "application"
-  module: "{module_code}"
-  test_type: "unit"
-
-Tests: service logic, CQRS handlers, FluentValidation
-```
-
-### API Tests (if not already scaffolded)
-```
-Call: mcp__smartstack__scaffold_tests
-  target_layer: "api"
-  module: "{module_code}"
-  test_type: "integration"
-
-Tests: controller endpoints, authorization, response DTOs
-```
-
-### Frontend Form Tests (if not already scaffolded)
-```
-For each module with create/edit pages:
-  → Generate EntityCreatePage.test.tsx and EntityEditPage.test.tsx
-  → Co-located next to the page component (NOT in __tests__/ folder)
-
-Required test coverage:
-  - Rendering: form renders with all expected fields
-  - Validation: required fields show errors on empty submit
-  - Submission: successful submit calls API and navigates back
-  - Pre-fill (edit only): entity data loaded into fields
-  - Navigation: back/cancel button calls navigate(-1)
-  - Error handling: API error displays error message
-
-Tools: Vitest + React Testing Library + @testing-library/user-event
-Mock API: vi.mock() or MSW — NEVER real API calls
-```
-
----
-
-## 4. Suggest Additional Scenarios
-
-```
-Call: mcp__smartstack__suggest_test_scenarios
-  module: "{module_code}"
-
-Review suggestions and add relevant test cases.
-```
-
----
-
-## 5. Analyze Coverage
-
-```
-Call: mcp__smartstack__analyze_test_coverage
-  module: "{module_code}"
-
-Target: >= 80% coverage
-If under 80%: identify uncovered paths, scaffold additional tests
-```
-
----
-
-## 6. Test Summary
-
-```
-**APEX SmartStack - Final Test Sweep**
-
-| Layer | Inline Tests | Final Sweep | Total |
-|-------|-------------|-------------|-------|
-| Domain | {count} (Layer 2) | {count} | {total} |
-| Application | {count} (Layer 2) | {count} | {total} |
-| API | {count} (Layer 2) | {count} | {total} |
-| Security | — | {count} | {count} |
-| Frontend Forms | {count} (Layer 3) | {count} | {total} |
-
-**Total tests:** {count}
-**Estimated coverage:** {%}
-```
-
----
-
-## 7. Save Output (if save_mode)
-
-Write to `{output_dir}/07-tests.md` with test scaffolding results.
-
----
-
-## NEXT STEP
-
-Load `steps/step-08-run-tests.md`

package/templates/skills/apex/steps/step-08-run-tests.md

@@ -1,135 +0,0 @@
----
-name: step-08-run-tests
-description: Run Final Test Sweep until 100% pass — fix CODE not tests
-model: opus
-prev_step: steps/step-07-tests.md
-next_step: COMPLETE
----
-
-# Step 8: Run Final Test Sweep
-
-**CRITICAL:** Fix the CODE, not the tests. Tests represent expected behavior.
-Target 100% pass for Sonar quality gate. If stuck after max retries, skip with TODO and escalate.
-
----
-
-## 1. Build
-
-```bash
-dotnet build --no-restore
-# Note: WSL bin\Debug cleanup handled by PostToolUse hook (wsl-dotnet-cleanup.sh)
-```
-
-**MUST PASS before running tests.**
-
----
-
-## 2. Run Full Test Suite (backend + frontend in parallel)
-
-```bash
-# Run backend and frontend tests in parallel when both exist
-WEB_DIR=$(find . -name "vitest.config.ts" -not -path "*/node_modules/*" -exec dirname {} \; 2>/dev/null | head -1)
-
-if [ -n "$WEB_DIR" ]; then
-  # Both exist → run in parallel
-  dotnet test --no-build --verbosity normal &
-  BACKEND_PID=$!
-  (cd "$WEB_DIR" && npm run test) &
-  FRONTEND_PID=$!
-  wait $BACKEND_PID; BACKEND_RC=$?
-  wait $FRONTEND_PID; FRONTEND_RC=$?
-  [ $BACKEND_RC -ne 0 ] || [ $FRONTEND_RC -ne 0 ] && echo "FAILURES DETECTED"
-else
-  # Backend only
-  dotnet test --no-build --verbosity normal
-fi
-```
-
----
-
-## 3. Analyze Results
-
-**If ALL PASS:** Skip to step 5 (Summary).
-
-**If FAILURES:** For each failing test:
-
-```markdown
-### Failing Test: {test_name}
-**File:** {test_file}:{line}
-**Error:** {error_message}
-
-**Root Cause Analysis:**
-- [ ] Code bug: implementation doesn't match expected behavior
-- [ ] Missing dependency: service/mock not configured
-- [ ] Data issue: test data setup incomplete
-```
-
----
-
-## 4. Fix Loop (max 5 iterations)
-
-> **Max 5 iterations (vs max 3 inline).** This catches cross-layer problems that inline tests missed.
-
-```
-WHILE tests failing AND iteration < 5:
-  1. Identify root cause (ALWAYS code bug, not test bug)
-  2. Fix the production CODE via appropriate skill/MCP
-  3. dotnet build --no-restore
-  4. dotnet test --no-build
-  5. Log result
-```
-
-**Rules:**
-- Fix CODE, never modify test assertions
-- Use the same skill/MCP that created the original code
-- After fix, rebuild before retesting
-- If stuck after 3 iterations on same test, ask user
-
-**If stuck (auto_mode = false):**
-
-```yaml
-questions:
-  - header: "Test Fix"
-    question: "Test is still failing after multiple attempts. How should we proceed?"
-    options:
-      - label: "Try alternative fix (Recommended)"
-        description: "Attempt a different approach to fix the code"
-      - label: "Skip this test"
-        description: "Mark as skipped with TODO comment"
-      - label: "Discuss the issue"
-        description: "Need help understanding the failure"
-    multiSelect: false
-```
-
----
-
-## 5. Typecheck (if applicable)
-
-> **Note:** Frontend tests already ran in parallel with backend tests in section 2.
-
-```bash
-npm run typecheck
-```
-
-**MUST PASS.** If frontend tests failed in section 2, apply the same fix loop as backend (fix CODE, not tests, max 5 iterations).
-
----
-
-## 6. Commit Tests
-
-```
-test({module}): security tests and coverage sweep
-```
-
----
-
-## 7. Save + Summary
-
-If save_mode: write to `{output_dir}/08-run-tests.md`.
-
-Display: total tests, passing count, iterations to green, fixes applied, coverage %.
-
-## 8. Route to Next Step
-
-If pr_mode: create PR via `gh pr create`, show URL.
-Otherwise: show final summary (task, context, files, quality checks, commits, next steps).

package/templates/skills/apex-verify/SKILL.md

@@ -1,110 +0,0 @@
----
-name: apex-verify
-description: |
-  Audit a generated SmartStack APEX application for common generation issues.
-  Use this skill when:
-  - Verifying a generated app has no missing CRUD pages or buttons
-  - Checking navigation seed data for reserved section codes in menu
-  - Auditing permission coverage (controllers vs Permissions.cs vs seed data)
-  - Validating route alignment (frontend vs backend vs seed)
-  - After /apex or /business-analyse-develop completes
-argument-hint: "[--scope=all|nav|crud|perm|route] [--fix]"
-model: sonnet
-allowed-tools: "Read, Grep, Glob, Bash, ToolSearch"
-entry_point: steps/step-00-init.md
----
-
-<objective>
-Audit a SmartStack application generated by /apex or /business-analyse-develop for common post-generation issues:
-
-1. **Navigation** — Reserved section codes (detail, create, edit) seeded as visible menu items
-2. **CRUD** — Missing page types, missing "New" buttons on list pages, incomplete componentRegistry
-3. **Permissions** — Cross-reference controllers, Permissions.cs, PermissionsSeedData, RolesSeedData
-4. **Routes** — Alignment between seed data routes, PageRegistry keys, and controller NavRoutes
-
-The audit is **read-only** by default. Use `--fix` to display actionable fix commands.
-</objective>
-
-<quick_start>
-
-```bash
-/apex-verify                    # Full audit (all 4 categories)
-/apex-verify --scope=nav        # Navigation menu audit only
-/apex-verify --scope=crud       # CRUD completeness audit only
-/apex-verify --scope=perm       # Permission cross-reference only
-/apex-verify --scope=route      # Route alignment only
-/apex-verify --fix              # Full audit + fix suggestions
-```
-
-</quick_start>
-
-<parameters>
-
-<flags>
-| Flag | Description |
-|------|-------------|
-| `--scope=all` | Audit all 4 categories (default) |
-| `--scope=nav` | Navigation audit only (reserved section codes) |
-| `--scope=crud` | CRUD completeness only (pages, buttons, registry) |
-| `--scope=perm` | Permission cross-reference only |
-| `--scope=route` | Route alignment only |
-| `--fix` | Display actionable fix commands for each finding |
-</flags>
-</parameters>
-
-<workflow>
-1. **Initialize** — Detect project structure, locate seed data, controllers, pages
-2. **Navigation Audit** — Check for reserved section codes seeded as menu items
-3. **CRUD Audit** — Verify 4 page types per section, create buttons, componentRegistry
-4. **Permission Audit** — Cross-reference controllers, Permissions.cs, seed data, roles
-5. **Route Audit** — Verify frontend/backend/seed alignment
-6. **Report** — Consolidated findings with severity and fix suggestions
-</workflow>
-
-<state_variables>
-| Variable | Type | Description |
-|----------|------|-------------|
-| `{scope}` | string | all, nav, crud, perm, route |
-| `{fix_mode}` | boolean | Display fix commands |
-| `{api_root}` | string | Path to backend src/ |
-| `{web_root}` | string | Path to frontend web/ |
-| `{seed_files}` | string[] | All *NavigationSeedData.cs files |
-| `{controller_files}` | string[] | All *Controller.cs files |
-| `{page_files}` | string[] | All *.tsx page files |
-| `{registry_file}` | string | componentRegistry.generated.ts path |
-| `{permissions_file}` | string | Permissions.cs path |
-| `{findings}` | object[] | All findings with ID, severity, message, fix |
-</state_variables>
-
-<entry_point>
-
-**FIRST ACTION:** Load `steps/step-00-init.md`
-
-</entry_point>
-
-<step_files>
-| Step | File | Purpose |
-|------|------|---------|
-| 00 | `steps/step-00-init.md` | Parse flags, detect project structure, resolve paths |
-| 01 | `steps/step-01-nav-audit.md` | Audit navigation seed data for reserved section codes |
-| 02 | `steps/step-02-crud-audit.md` | Audit CRUD completeness (pages, buttons, registry) |
-| 03 | `steps/step-03-perm-audit.md` | Audit permissions cross-reference |
-| 04 | `steps/step-04-route-audit.md` | Audit route alignment |
-| 05 | `steps/step-05-report.md` | Generate consolidated report |
-</step_files>
-
-<execution_rules>
-- **Read-only** — This skill NEVER modifies code (unless --fix is used, and even then only displays commands)
-- **Evidence-based** — Every finding must reference a specific file:line
-- **Comprehensive** — Check ALL files, not a sample
-- **Actionable** — Every finding must include a fix suggestion
-- **Scoped** — Only run audits matching {scope}
-- **SmartStack-aware** — Respect SmartStack conventions (PageRegistry, DynamicRouter, NavRoute, SeedData patterns)
-</execution_rules>
-
-<success_criteria>
-- All applicable audits completed with zero false positives
-- Each finding has: ID, severity (BLOCKING/CRITICAL/WARNING), file:line, description, fix
-- Consolidated report with category totals
-- Executive summary with pass/fail per category
-</success_criteria>

package/templates/skills/apex-verify/references/audit-rules.md

@@ -1,50 +0,0 @@
-# Audit Rules Reference
-
-> Complete codification of all audit rules used by /apex-verify.
-> Each rule has a unique ID, severity, description, detection method, and fix pattern.
-
----
-
-## Navigation Rules (NAV-xxx)
-
-| ID | Severity | Description | Detection | Fix |
-|----|----------|-------------|-----------|-----|
-| NAV-001 | BLOCKING | Reserved section code (detail/create/edit) seeded as visible menu item | Grep NavigationSeedData for `Code = "detail"` / `"create"` / `"edit"` without `IsActive = false` | Remove section from seed data, or set `IsActive = false` |
-| NAV-002 | BLOCKING | Section code contains "-detail" seeded as menu item | Grep for codes matching `*-detail` pattern | Remove section entry — detail is resolved as /:id under parent section |
-| NAV-003 | BLOCKING | Route contains forbidden segment (/detail/, /create/, /edit/) | Grep seed data routes for `/detail/`, `/create/`, `/edit/` segments | Use convention routes: /:id, /create, /:id/edit (no explicit segments) |
-| NAV-004 | WARNING | Reserved section code with IsActive = false | Code is reserved but hidden from menu | Acceptable — remove if unnecessary |
-| NAV-005 | WARNING | Section code not in kebab-case | Regex check: `^[a-z][a-z0-9]*(-[a-z0-9]+)*$` | Rename to kebab-case |
-
-## CRUD Rules (CRUD-xxx)
-
-| ID | Severity | Description | Detection | Fix |
-|----|----------|-------------|-----------|-----|
-| CRUD-001 | BLOCKING | ListPage missing for entity | Glob for `*ListPage.tsx` matching entity | Generate via `/ui-components` |
-| CRUD-002 | BLOCKING | DetailPage missing for entity | Glob for `*DetailPage.tsx` matching entity | Generate via `/ui-components` |
-| CRUD-003 | BLOCKING | Create/Edit page missing for entity | Glob for `*CreatePage.tsx`, `*EditPage.tsx`, `*FormPage.tsx` | Generate via `/ui-components` |
-| CRUD-004 | BLOCKING | ListPage has no Create/New button | Grep for `navigate.*create` or `Link.*create` | Add navigate('create') button in page header |
-| CRUD-005 | BLOCKING | Page exists but not registered in componentRegistry | Read registry, check for matching key | Add `PageRegistry.register()` call or run `scaffold_routes` |
-| CRUD-006 | WARNING | Registry entry has no matching page file | Read registry, check file existence | Remove orphan registry entry or create the page |
-| CRUD-007 | WARNING | Form page has no submit handler | Grep for `onSubmit`, `handleSubmit`, `api.create`, `api.post` | Implement form submission logic |
-
-## Permission Rules (PERM-xxx)
-
-| ID | Severity | Description | Detection | Fix |
-|----|----------|-------------|-----------|-----|
-| PERM-001 | BLOCKING | Controller uses permission not defined in Permissions.cs | Cross-ref `[RequirePermission]` vs `Permissions.cs` | Add missing constant to Permissions.cs |
-| PERM-002 | BLOCKING | Permissions.cs constant not seeded in PermissionsSeedData | Cross-ref Permissions.cs paths vs seed data | Add `HasData()` entry in PermissionsSeedData |
-| PERM-003 | CRITICAL | Seeded permission has no role mapping | Cross-ref seed data vs RolesSeedData | Add role-permission mapping in RolesSeedData |
-| PERM-004 | CRITICAL | Admin role missing wildcard for module | Check admin role has `*.module.*` wildcard | Add wildcard permission for admin role |
-| PERM-005 | WARNING | Permission path segments not in kebab-case | Regex check on permission paths | Rename to kebab-case |
-| PERM-006 | WARNING | Viewer role has write permission | Check viewer role permissions | Remove write permissions from viewer role |
-| PERM-007 | WARNING | Permission defined but unused by any controller | Cross-ref Permissions.cs vs controllers | Remove dead permission or add controller usage |
-
-## Route Rules (ROUTE-xxx)
-
-| ID | Severity | Description | Detection | Fix |
-|----|----------|-------------|-----------|-----|
-| ROUTE-001 | BLOCKING | Seed data route has no matching PageRegistry entry | Cross-ref seed routes vs registry keys | Add PageRegistry.register() or run scaffold_routes |
-| ROUTE-002 | WARNING | Controller NavRoute has no matching seed data entry | Cross-ref NavRoute attrs vs seed routes | Add navigation seed data entry |
-| ROUTE-003 | BLOCKING | PageRegistry references non-existent page file | Read registry imports, check file existence | Create page or fix import path |
-| ROUTE-004 | WARNING | Orphan page — exists on disk but not registered | Cross-ref page files vs registry | Register page or delete orphan |
-| ROUTE-005 | WARNING | Implicit route (detail/create/edit) not resolvable | Check registry + DynamicRouter conventions | Register in componentRegistry |

package/templates/skills/apex-verify/steps/step-00-init.md

@@ -1,119 +0,0 @@
----
-name: step-00-init
-description: Parse flags, detect project structure, resolve paths for APEX verification
-next_step: steps/step-01-nav-audit.md
----
-
-# Step 0: Initialization
-
-## MANDATORY EXECUTION RULES:
-- NEVER skip project detection
-- ALWAYS resolve actual filesystem paths before proceeding
-- YOU ARE AN INITIALIZER, not a scanner
-
-## YOUR TASK:
-Parse the flags, detect the SmartStack project structure, and locate all files needed for auditing.
-
----
-
-## EXECUTION SEQUENCE:
-
-### 1. Parse Input
-
-```
-Arguments from $ARGUMENTS:
-  --scope=nav       -> {scope} = "nav"
-  --scope=crud      -> {scope} = "crud"
-  --scope=perm      -> {scope} = "perm"
-  --scope=route     -> {scope} = "route"
-  --scope=all       -> {scope} = "all" (default)
-  --fix             -> {fix_mode} = true (default: false)
-
-No arguments -> {scope} = "all", {fix_mode} = false
-```
-
-### 2. Detect Project Structure
-
-Use Glob to locate all required files:
-
-```
-Backend paths:
-  {api_root} = src/*/Api/ or src/*.Api/
-  Verify: *.sln exists at project root
-
-Frontend paths:
-  {web_root} = web/*/src/ or web/*-web/src/
-  Verify: package.json exists in web/*/
-
-Key files to locate:
-  {seed_files}:
-    Glob "**/Seeding/Data/**/*NavigationSeedData.cs"
-    Glob "**/Seeding/Data/**/*NavigationModuleSeedData.cs"
-
-  {controller_files}:
-    Glob "**/Controllers/**/*Controller.cs"
-
-  {page_files}:
-    Glob "web/**/src/pages/**/*Page.tsx"
-
-  {registry_file}:
-    Glob "**/componentRegistry.generated.ts" or "**/componentRegistry*.ts"
-
-  {permissions_file}:
-    Glob "**/Authorization/Permissions.cs" or "**/Common/Authorization/Permissions.cs"
-
-  {perm_seed_files}:
-    Glob "**/Seeding/Data/**/*PermissionsSeedData.cs"
-    Glob "**/Seeding/Data/**/*SeedDataProvider.cs"
-
-  {role_seed_files}:
-    Glob "**/Seeding/Data/**/*RolesSeedData.cs"
-```
-
-### 3. Validate Paths
-
-For each required file set:
-- If empty: record as CRITICAL warning (cannot audit that category)
-- If found: record paths for next steps
-
-### 4. Show Summary and Proceed
-
-Display:
-
-```
-APEX Verification — Initialization
-====================================
-
-| Setting | Value |
-|---------|-------|
-| Scope | {scope} |
-| Fix mode | {fix_mode} |
-
-Project Structure:
-| Category | Files Found |
-|----------|-------------|
-| Navigation seed data | {count} files |
-| Controllers | {count} files |
-| Frontend pages | {count} files |
-| Component registry | {found/MISSING} |
-| Permissions.cs | {found/MISSING} |
-| Permission seed data | {count} files |
-| Role seed data | {count} files |
-
--> Proceeding to audits...
-```
-
----
-
-## SUCCESS METRICS:
-- Scope and flags correctly parsed
-- All filesystem paths resolved and validated
-- Summary displayed with file counts
-- Ready for audit steps
-
-## NEXT STEP:
-Based on {scope}:
-- `all` or `nav` → proceed to `./step-01-nav-audit.md`
-- `crud` → skip to `./step-02-crud-audit.md`
-- `perm` → skip to `./step-03-perm-audit.md`
-- `route` → skip to `./step-04-route-audit.md`