data-and-database-coupling-evaluator

Evaluate ORM configurations, schema changes, and data access layers for shared-table coupling, ORM leakage, and distributed data joins. Use when reviewing schema migrations, repository classes, or modernization work toward decoupled data ownership.

Data and Database Coupling Evaluator

Analyze data access boundaries to detect model coupling and shared-table integration.

When to Use

  • PRs with schema migrations
  • ORM entity or repository changes
  • Data access layer refactors
  • Monolith-to-microservice decomposition reviews

Scope

  • Scans: ORM entities, SQL migration scripts, DAO/repository classes, and SQL query strings that reveal table ownership and schema coupling.
  • Exclusions: analytics/read-model joins that are explicitly documented, generated code, and runtime-only database issues without static evidence.
  • Fallback: if schema ownership is ambiguous, fall back to the shared conventions order and downgrade to warning with low confidence.
  • Routing: defer bounded-context translation concerns to Domain Boundary and Language Integrity Evaluator and distributed workflow ownership to Distributed Workflow and Integration Analyzer.

Workflow

  1. Detect shared table coupling.
    • Map modules or bounded contexts to the tables they read or write.
    • Flag multiple writers as high severity.
    • Flag cross-module reads from another module's table as moderate severity.
  2. Detect ORM leakage.
    • Check whether ORM entities or annotations cross into API or domain boundaries.
    • Prefer DTOs and isolated domain entities.
  3. Detect distributed data joins.
    • Identify cross-module aggregation patterns, app-level joins, and cross-boundary N+1 risks.
    • Prefer dedicated read models or analytics stores when needed.
  4. Report findings.
    • List shared tables and operation types.
    • Include file paths for ORM leakage.
    • Format output for CI/CD consumption.

Rules

  • Use static analysis only.
  • Do not connect to a live database.
  • Keep analysis deterministic and language-agnostic.
  • Treat shared writes as the most severe coupling signal.

Shared Conventions

  • Follow the repository-wide conventions in shared-skill-conventions.
  • Use the canonical output contract, severity levels, confidence rubric, routing rules, and false-positive downgrades defined there.
  • If this skill is not the closest match, defer to the routing guidance in the shared conventions file.

Output

Return:

  • Shared tables with module-level access details
  • ORM leakage findings
  • Distributed join and N+1 risks
  • CI-ready pass or fail output