ezorm-write-yaml-mongo_skill

Overview

This skill generates ezorm YAML schema files specifically for MongoDB collections based on a provided domain model or update request. It produces only valid YAML documents, focused on collections, fields, indexes, embeds, and relations for mongo. Use it when you need schema-first YAML for ezorm without any Go code output.

How this skill works

I parse the domain model and map entities to ezorm YAML using mongo-specific types and flags. I decide between embedded documents (embed: true) and separate collections, add indexes/uniques/primary as requested, and populate collection metadata (table and dbname). The output is minimal, consistent, and adheres to ezorm mongo conventions.

When to use it

• Defining new MongoDB collections and their fields in ezorm YAML
• Updating indexes, unique constraints, or flags for mongo entities
• Modeling embedded documents versus separate collections
• Adding relations that use ObjectID references for mongo
• Generating YAML for validation or CI schema checks

Best practices

• Describe entities and relationships clearly; I will ask one clarifying question only if essential
• Prefer embeds for tightly coupled subdocuments and separate collections for many-to-many or large collections
• Specify storage names only when they differ from logical names so I avoid unnecessary bson/json tags
• Include required index and unique constraints in the request to ensure correct flags
• Provide dbname when it should differ from 'default' to avoid assumptions

Example use cases

• Create ezorm YAML for a blog app: posts, authors, comments with comments embedded in posts
• Add a unique email index and login count index to a Users collection
• Migrate a product catalog: embed pricing details or move them to a separate collection
• Define relationships using ObjectID fields for orders referencing customers
• Update schema to add a compound index for query sorting and filtering