AutoBE
    Preparing search index...

    Interface AutoBeDatabaseComponentTableDesign

    Lightweight table design used before the full Prisma model is generated.

    This structure identifies table concepts only. Field lists, indexes, relations, constraints, and column-level modeling belong to the later database schema generation step.

    Michael

    interface AutoBeDatabaseComponentTableDesign {
        description: string;
        name: string & SnakeCasePattern;
        stance:
            | "actor"
            | "primary"
            | "subsidiary"
            | "snapshot"
            | "material"
            | "session";
    }
    Index

    Properties

    description: string

    Concise business concept of this table. One or two sentences maximum. MUST be written in English. State the stored concept, owner, row grain, lifecycle trigger, and concept-level retained fact categories so later review agents can compare this table with similar table concepts.

    Do not list columns, indexes, relation definitions, or schema constraints here. If a current/display value is derived from source rows in another table, describe that source relationship instead of implying this table stores the value.

    name: string & SnakeCasePattern

    Final table name for this table concept.

    Use a plural, snake_case, prefix-aware table name that represents one stored concept. Apply the configured database prefix when one exists and avoid duplicate prefixes. The name must identify the table itself, not a column, field, relation, index, or constraint.

    Examples: shopping_sales, shopping_sale_snapshots, shopping_sale_snapshot_units.

    stance: "actor" | "primary" | "subsidiary" | "snapshot" | "material" | "session"

    Architectural role of this table, guiding API endpoint generation.

    • "primary": Core entity users manage independently (full CRUD APIs). Use when users need to create, search, or manage entities outside their parent context.
    • "actor": Authenticated user type with its own identity, credentials, and auth flow. Generates auth endpoints.
    • "session": Login session table belonging to exactly one actor. Append-only audit trail, managed via auth flows.
    • "subsidiary": Supporting entity managed through its parent, rarely needs standalone endpoints.
    • "snapshot": Point-in-time versioning record, typically append-only and read-only from user perspective.
    • "material": Read-only materialized query projection. The table name must start with mv_, and normal CRUD endpoints are not generated.