@lenne.tech/cli 1.0.0 → 1.0.1

package/build/templates/claude-skills/nest-server-generator/reference.md

@@ -1,6 +1,6 @@
 ---
 name: nest-server-generator-reference
-version: 1.0.0
+version: 1.0.1
 description: Quick reference for ALL NestJS server development - from simple single commands to complex structure generation
 ---
 
@@ -63,7 +63,13 @@ Use this skill for **ANY** NestJS/nest-server work, no matter how simple or comp
 ☐ 4. Create all Objects
 ☐ 5. Create all Modules (dependency order)
 ☐ 6. Handle inheritance (manual edits)
-☐ 7. Update all descriptions (ENGLISH (DEUTSCH))
+☐ 7. Update ALL descriptions EVERYWHERE (CRITICAL!)
+    ☐ 7.1. Extract ALL user comments (after //) from specification
+    ☐ 7.2. Format descriptions: ENGLISH (DEUTSCH)
+    ☐ 7.3. Apply to ALL Module files (Model, CreateInput, UpdateInput)
+    ☐ 7.4. Apply to ALL SubObject files (Object, CreateInput, UpdateInput)
+    ☐ 7.5. Add to ALL class decorators (@ObjectType, @InputType)
+    ☐ 7.6. Verify consistency (same property = same description)
 ☐ 8. Alphabetize all properties
 ☐ 9. Create enum files
 ☐ 10. Create API tests
@@ -140,20 +146,57 @@ lt server addProp --type Module --element <Name> \
 
 ## Description Format
 
+**⚠️ CRITICAL:** Always extract descriptions from user comments (after `//`) and apply EVERYWHERE!
+
 **Rule**: `"ENGLISH_DESCRIPTION (DEUTSCHE_BESCHREIBUNG)"`
 
 ### Processing
 
-| Input Comment | Output Description |
-|---------------|-------------------|
-| `// Street name` | `'Street name'` |
-| `// Straße` | `'Street (Straße)'` |
-| (no comment) | Create meaningful English description |
-
-### Apply To
-- Model property `@UnifiedField({ description: '...' })`
-- Input property `@UnifiedField({ description: '...' })`
-- Output property `@UnifiedField({ description: '...' })`
+| Input Comment | Language | Output Description |
+|---------------|----------|-------------------|
+| `// Product name` | English | `'Product name'` |
+| `// Produktname` | German | `'Product name (Produktname)'` |
+| `// Street name` | English | `'Street name'` |
+| `// Straße` | German | `'Street (Straße)'` |
+| `// Postleizahl` (typo) | German | `'Postal code (Postleitzahl)'` (corrected) |
+| (no comment) | - | Create meaningful English description |
+
+**⚠️ Preserve Original Wording:**
+- ✅ Fix typos: `Postleizahl` → `Postleitzahl`, `Starße` → `Straße`
+- ❌ DON'T rephrase: `Straße` → `Straßenname` (NO!)
+- ❌ DON'T expand: `Produkt` → `Produktbezeichnung` (NO!)
+- **Reason:** User comments may be predefined terms referenced by external systems
+
+### Apply To ALL Files
+
+**For EVERY Module property** (3 files):
+1. `<module>.model.ts` → Property `@UnifiedField({ description: '...' })`
+2. `inputs/<module>-create.input.ts` → Property `@UnifiedField({ description: '...' })`
+3. `inputs/<module>.input.ts` → Property `@UnifiedField({ description: '...' })`
+
+**For EVERY SubObject property** (3 files):
+1. `objects/<object>/<object>.object.ts` → Property `@UnifiedField({ description: '...' })`
+2. `objects/<object>/<object>-create.input.ts` → Property `@UnifiedField({ description: '...' })`
+3. `objects/<object>/<object>.input.ts` → Property `@UnifiedField({ description: '...' })`
+
+**For class decorators**:
+- `@ObjectType({ description: '...' })` on Models and Objects
+- `@InputType({ description: '...' })` on all Input classes
+
+### Common Mistakes
+
+❌ **WRONG:** Descriptions only in Model, missing in Inputs
+❌ **WRONG:** German-only descriptions without English translation
+❌ **WRONG:** Inconsistent descriptions (different in Model vs Input)
+❌ **WRONG:** Ignoring user-provided comments from specification
+❌ **WRONG:** Changing wording: `Straße` → `Straßenname` (rephrased!)
+❌ **WRONG:** Expanding terms: `Produkt` → `Produktbezeichnung` (added word!)
+
+✅ **CORRECT:** Same description in ALL 3 files (Model, CreateInput, UpdateInput)
+✅ **CORRECT:** Format `ENGLISH (DEUTSCH)` for German comments
+✅ **CORRECT:** All user comments extracted and applied
+✅ **CORRECT:** Fix typos only, preserve original wording: `Postleizahl` → `Postleitzahl`
+✅ **CORRECT:** Keep exact terms: `Straße` → `Street (Straße)` (not "Street name"!)
 
 ## Inheritance Handling
 
@@ -326,7 +369,18 @@ Final checks before completing:
 ☐ All Objects created
 ☐ All Modules created
 ☐ Properties in alphabetical order
-☐ Descriptions: "ENGLISH (DEUTSCH)" format
+☐ DESCRIPTIONS - CRITICAL (check ALL):
+  ☐ User comments extracted from specification
+  ☐ German descriptions → ENGLISH (DEUTSCH) format
+  ☐ English descriptions → kept as-is
+  ☐ Module Models have descriptions
+  ☐ Module CreateInputs have SAME descriptions
+  ☐ Module UpdateInputs have SAME descriptions
+  ☐ SubObjects have descriptions
+  ☐ SubObject CreateInputs have SAME descriptions
+  ☐ SubObject UpdateInputs have SAME descriptions
+  ☐ @ObjectType/@InputType decorators have descriptions
+  ☐ NO inconsistencies between files
 ☐ Inheritance correctly implemented
 ☐ CreateInputs have all required fields (parent + model)
 ☐ Enum files created in src/server/common/enums/

package/build/templates/claude-skills/nest-server-generator/security-rules.md

@@ -0,0 +1,358 @@
+---
+name: nest-server-generator-security-rules
+version: 1.0.0
+description: Critical security and test coverage rules for NestJS development
+---
+
+# 🚨 CRITICAL SECURITY RULES
+
+**Before you start ANY work, understand these NON-NEGOTIABLE rules.**
+
+---
+
+## ⛔ NEVER Do This
+
+1. **NEVER remove or weaken `@Restricted()` decorators** to make tests pass
+2. **NEVER change `@Roles()` decorators** to more permissive roles for test convenience
+3. **NEVER modify `securityCheck()` logic** to bypass security in tests
+4. **NEVER remove class-level `@Restricted(RoleEnum.ADMIN)`** - it's a security fallback
+
+---
+
+## ✅ ALWAYS Do This
+
+1. **ALWAYS analyze permissions BEFORE writing tests** (Controller, Model, Service layers)
+2. **ALWAYS test with the LEAST privileged user** who is authorized
+3. **ALWAYS create appropriate test users** for each permission level
+4. **ALWAYS adapt tests to security requirements**, never the other way around
+5. **ALWAYS ask developer for approval** before changing ANY security decorator
+6. **ALWAYS aim for maximum test coverage** (80-100% depending on criticality)
+
+---
+
+## 🔑 Permission Hierarchy (Specific Overrides General)
+
+```typescript
+@Restricted(RoleEnum.ADMIN)  // ← FALLBACK: DO NOT REMOVE
+export class ProductController {
+  @Roles(RoleEnum.S_USER)    // ← SPECIFIC: This method is more open
+  async createProduct() { }   // ← S_USER can access (specific wins)
+
+  async secretMethod() { }    // ← ADMIN only (fallback applies)
+}
+```
+
+**Why class-level `@Restricted(ADMIN)` MUST stay:**
+- If someone forgets `@Roles()` on a new method → it's secure by default
+- Shows the class is security-sensitive
+- Fail-safe protection
+
+---
+
+## Rule 1: NEVER Weaken Security for Test Convenience
+
+### ❌ ABSOLUTELY FORBIDDEN
+
+```typescript
+// BEFORE (secure):
+@Restricted(RoleEnum.ADMIN)
+export class ProductController {
+  @Roles(RoleEnum.S_USER)
+  async createProduct() { ... }
+}
+
+// AFTER (FORBIDDEN - security weakened!):
+// @Restricted(RoleEnum.ADMIN)  ← NEVER remove this!
+export class ProductController {
+  @Roles(RoleEnum.S_USER)
+  async createProduct() { ... }
+}
+```
+
+### 🚨 CRITICAL RULE
+
+- **NEVER remove or weaken `@Restricted()` decorators** on Controllers, Resolvers, Models, or Objects
+- **NEVER change `@Roles()` decorators** to more permissive roles just to make tests pass
+- **NEVER modify `securityCheck()` logic** to bypass security for testing
+
+### If tests fail due to permissions
+
+1. ✅ **CORRECT**: Adjust the test to use the appropriate user/token
+2. ✅ **CORRECT**: Create test users with the required roles
+3. ❌ **WRONG**: Weaken security to make tests pass
+
+### Any security changes MUST
+
+- Be discussed with the developer FIRST
+- Have a solid business justification
+- Be explicitly approved by the developer
+- Be documented with the reason
+
+---
+
+## Rule 2: Understanding Permission Hierarchy
+
+### ⭐ Key Concept: Specific Overrides General
+
+The `@Restricted()` decorator on a class acts as a **security fallback** - if a method/property doesn't specify permissions, it inherits the class-level restriction. This is a **security-by-default** pattern.
+
+### Example - Controller/Resolver
+
+```typescript
+@Restricted(RoleEnum.ADMIN)  // ← FALLBACK: Protects everything by default
+export class ProductController {
+
+  @Roles(RoleEnum.S_EVERYONE)  // ← SPECIFIC: This method is MORE open
+  async getPublicProducts() {
+    // Anyone can access this (specific @Roles wins)
+  }
+
+  @Roles(RoleEnum.S_USER)  // ← SPECIFIC: Logged-in users
+  async createProduct() {
+    // S_USER can access (specific wins over fallback)
+  }
+
+  async deleteProduct() {
+    // ADMIN ONLY (no specific decorator, fallback applies)
+  }
+}
+```
+
+### Example - Model
+
+```typescript
+@Restricted(RoleEnum.ADMIN)  // ← FALLBACK
+export class Product {
+
+  @Roles(RoleEnum.S_EVERYONE)  // ← SPECIFIC
+  @UnifiedField({ description: 'Product name' })
+  name: string;  // Everyone can read this
+
+  @UnifiedField({ description: 'Internal cost' })
+  cost: number;  // ADMIN ONLY (fallback applies)
+}
+```
+
+---
+
+## Rule 3: Adapt Tests to Security, Not Vice Versa
+
+### ❌ WRONG Approach
+
+```typescript
+// Test fails because user isn't admin
+it('should create product', async () => {
+  const result = await request(app)
+    .post('/products')
+    .set('Authorization', regularUserToken)  // Not an admin!
+    .send(productData);
+
+  expect(result.status).toBe(201);  // Fails with 403
+});
+
+// ❌ WRONG FIX: Removing @Restricted from controller
+// @Restricted(RoleEnum.ADMIN)  ← NEVER DO THIS!
+```
+
+### ✅ CORRECT Approach
+
+```typescript
+// Analyze first: Who is allowed to create products?
+// Answer: ADMIN only (based on @Restricted on controller)
+
+// Create admin test user
+let adminToken: string;
+
+beforeAll(async () => {
+  const admin = await createTestUser({ roles: [RoleEnum.ADMIN] });
+  adminToken = admin.token;
+});
+
+it('should create product as admin', async () => {
+  const result = await request(app)
+    .post('/products')
+    .set('Authorization', adminToken)  // ✅ Use admin token
+    .send(productData);
+
+  expect(result.status).toBe(201);  // ✅ Passes
+});
+
+it('should reject product creation for regular user', async () => {
+  const result = await request(app)
+    .post('/products')
+    .set('Authorization', regularUserToken)
+    .send(productData);
+
+  expect(result.status).toBe(403);  // ✅ Test security works!
+});
+```
+
+---
+
+## Rule 4: Test with Least Privileged User
+
+**Always test with the LEAST privileged user who is authorized to perform the action.**
+
+### ❌ WRONG
+
+```typescript
+// Method allows S_USER, but testing with ADMIN
+@Roles(RoleEnum.S_USER)
+async getProducts() { }
+
+it('should get products', async () => {
+  const result = await request(app)
+    .get('/products')
+    .set('Authorization', adminToken);  // ❌ Over-privileged!
+});
+```
+
+### ✅ CORRECT
+
+```typescript
+@Roles(RoleEnum.S_USER)
+async getProducts() { }
+
+it('should get products as regular user', async () => {
+  const result = await request(app)
+    .get('/products')
+    .set('Authorization', regularUserToken);  // ✅ Least privilege
+});
+```
+
+**Why this matters:**
+- Tests might pass with ADMIN but fail with S_USER
+- You won't catch permission bugs
+- False confidence in security
+
+---
+
+## Rule 5: Create Appropriate Test Users
+
+**Create test users for EACH permission level you need to test.**
+
+### Example Test Setup
+
+```typescript
+describe('ProductController', () => {
+  let adminToken: string;
+  let userToken: string;
+  let everyoneToken: string;
+
+  beforeAll(async () => {
+    // Create admin user
+    const admin = await createTestUser({
+      roles: [RoleEnum.ADMIN]
+    });
+    adminToken = admin.token;
+
+    // Create regular user
+    const user = await createTestUser({
+      roles: [RoleEnum.S_USER]
+    });
+    userToken = user.token;
+
+    // Create unauthenticated scenario
+    const guest = await createTestUser({
+      roles: [RoleEnum.S_EVERYONE]
+    });
+    everyoneToken = guest.token;
+  });
+
+  it('admin can delete products', async () => {
+    // Use adminToken
+  });
+
+  it('regular user can create products', async () => {
+    // Use userToken
+  });
+
+  it('everyone can view products', async () => {
+    // Use everyoneToken or no token
+  });
+
+  it('regular user cannot delete products', async () => {
+    // Use userToken, expect 403
+  });
+});
+```
+
+---
+
+## Rule 6: Comprehensive Test Coverage
+
+**Aim for 80-100% test coverage depending on criticality:**
+
+- **High criticality** (payments, user data, admin functions): 95-100%
+- **Medium criticality** (business logic, CRUD): 80-90%
+- **Low criticality** (utilities, formatters): 70-80%
+
+### What to Test
+
+**For each endpoint/method:**
+
+1. ✅ Happy path (authorized user, valid data)
+2. ✅ Permission denied (unauthorized user)
+3. ✅ Validation errors (invalid input)
+4. ✅ Edge cases (empty data, boundaries)
+5. ✅ Error handling (server errors, missing resources)
+
+### Example Comprehensive Tests
+
+```typescript
+describe('createProduct', () => {
+  it('should create product with admin user', async () => {
+    // Happy path
+  });
+
+  it('should reject creation by regular user', async () => {
+    // Permission test
+  });
+
+  it('should reject invalid product data', async () => {
+    // Validation test
+  });
+
+  it('should reject duplicate product name', async () => {
+    // Business rule test
+  });
+
+  it('should handle missing required fields', async () => {
+    // Edge case
+  });
+});
+```
+
+---
+
+## Quick Security Checklist
+
+Before completing ANY task:
+
+- [ ] **All @Restricted decorators preserved**
+- [ ] **@Roles decorators NOT made more permissive**
+- [ ] **Tests use appropriate user roles**
+- [ ] **Test users created for each permission level**
+- [ ] **Least privileged user tested**
+- [ ] **Permission denial tested (403 responses)**
+- [ ] **No securityCheck() logic bypassed**
+- [ ] **Test coverage ≥ 80%**
+- [ ] **All edge cases covered**
+
+---
+
+## Security Decision Protocol
+
+**When you encounter a security-related decision:**
+
+1. **STOP** - Don't make the change immediately
+2. **ANALYZE** - Why does the current security exist?
+3. **ASK** - Consult the developer before changing
+4. **DOCUMENT** - If approved, document the reason
+5. **TEST** - Ensure security still works after change
+
+**Remember:**
+- **Security > Convenience**
+- **Better to over-restrict than under-restrict**
+- **Always preserve existing security mechanisms**
+- **When in doubt, ask the developer**