| name | mongodb-mean |
| description | MongoDB database architecture for MEAN stack apps. Use when designing collections, schemas, indexes, relationships, and data modeling with Mongoose ODM. |
Mongodb Mean
Purpose
Design MongoDB database architectures for MEAN stack applications. Produce Mongoose schema definitions, index strategies, relationship patterns, and data modeling decisions that backend developers implement directly.
Key Patterns
Schema Design Principles
- Embed vs Reference: Embed data accessed together (comments in a post). Reference data accessed independently (user profiles from posts). If embedded array can grow unbounded, reference instead.
- Schema-first with Mongoose: Always define explicit schemas. Use Mongoose schema types, validators, and virtuals.
- Denormalize for reads: Duplicate frequently-read fields (e.g., author.name in posts) to avoid joins.
Standard Mongoose Patterns
- Always include
{ timestamps: true } — adds createdAt, updatedAt
- Soft delete pattern:
{ isDeleted: Boolean, deletedAt: Date }
- Pagination-ready: always index on createdAt
- Enum validation:
{ role: { type: String, enum: ['user', 'admin'], default: 'user' } }
- Reference pattern:
{ author: { type: mongoose.Schema.Types.ObjectId, ref: 'User', required: true } }
Index Strategy
- Every query field gets an index
- Compound indexes for multi-field queries:
{ userId: 1, createdAt: -1 }
- Unique indexes for natural keys:
{ email: 1 } with unique: true
- Text indexes for search:
{ title: 'text', body: 'text' }
Relationship Patterns
- One-to-Few (<=20): Embed directly (addresses in user)
- One-to-Many (20-1000): Array of references (posts by user)
- One-to-Millions: Parent reference in child (log entries referencing server)
- Many-to-Many: Array of references on both sides, or junction collection
Conventions
- Collection names: lowercase plural (users, posts, comments)
- Field names: camelCase (firstName, createdAt, isActive)
- Always include timestamps: true in every schema
- Always define toJSON transform to remove __v and convert _id to id
- Define indexes in schema file, not in application code
- Use Mongoose middleware (pre/post hooks) for cross-cutting concerns
- Validate at schema level (required, min/max, enum, match)
Knowledge Strategy
- Patterns to capture: Schema designs that worked well, index strategies for common queries, relationship patterns
- Examples to collect: Complete schema files for common entities (User, Post, Comment, Product, Order)
- Update permission: Agents may freely add/update files in
references/. Changes to SKILL.md or scripts/ require user approval.
