Business Invariants & Rule Ownership Separate transport validation, business rules, authorization, and database constraints
Phase 14 - Service & Business Logic Architecture Volume 04 - Professional Backend
v
v
Phase 14 turns CRUD-oriented service methods into explicit application rules and workflows while keeping infrastructure simple.
Invariant: a rule that must remain true A business invariant is a statement the application should protect across valid operations. It is different from a TypeScript type, a request shape, and often different from a database constraint.
title is a string <= 200 chars Zod boundary
projectId references an existing row PostgreSQL FK + service-friendly 404
assignee belongs to Project workspace TasksService
workspaceId + userId unique PostgreSQL unique constraint
caller may edit this Task authorization layer later
Why some rules cannot be expressed as simple database constraints The assignee rule crosses Task -> Project -> Workspace and Membership. PostgreSQL can ensure assigneeId references a real User, but the current model does not encode the stronger rule that the User must also belong to the Task workspace. The application service therefore checks that invariant.
v
v
v
v
A valid foreign key is necessary but not sufficient for this business rule.
Validation layers complement one another Layer Question
Zod Is assigneeId shaped like a UUID?
TasksService Is that User a member of this Workspace?
PostgreSQL Does that User row actually exist?
Request schema validation should stay deterministic and focused on shape/format. Database-dependent business rules belong in application logic, not inside the transport schema.
Authorization decides what an authenticated identity may do to a specific resource. UI hiding is never a substitute for server-side enforcement.
Restate the goal of “Business Invariants and Rule Ownership” as one observable behavior. Write the success signal before touching code.
Trace the boundary map: trusted identity → requested action → resource scope → policy decision → allow/deny. Mark the one boundary this lesson changes.
Inspect the current repository state and record the exact files, route, table, or runtime involved. Avoid creating an abstraction before you can name the existing owner.
Implement the smallest vertical slice. Add one coherent change, run the narrowest verification command, and read the output before continuing.
Connect the slice to TaskFlow’s phase artifact: Refactored task service with explicit policies/invariants and thin controllers.. Verify behavior from the outside as a user or API client would.
Create one controlled failure that is safe to reproduce. Predict the symptom first, capture evidence, locate the earliest broken boundary, then fix only that cause.
Close the reference and rebuild or explain the core change from memory. Record what you would monitor or test before trusting the same change in production.
const canEdit =
user.role === "ADMIN" ||
(user.role === "MEMBER" && task.ownerId === user.id);
if (!canEdit) throw new ForbiddenException();Do not copy this as a finished answer. Predict what each boundary does, type the smallest useful slice, run it, then compare the result with your prediction.
Trace verified identity, role, resource owner/workspace, and policy inputs separately.
Add server-side resource authorization; UI visibility is only experience, never enforcement.
Without copying the final source, create a second small scenario that applies “Business Invariants and Rule Ownership” to a different TaskFlow resource or input. Write the acceptance criteria first, implement it, and provide the evidence that each criterion is satisfied.
Move HTTP-only concerns back to controller/boundary and persistence mechanics behind a focused collaborator.
Use the smallest change that demonstrates this lesson. Stop after it works and capture evidence.Business rules duplicated in controllers and frontend
Predict the symptom before causing the failure. Then locate the earliest boundary where reality diverges from expectation.Why does this responsibility live here, what input does it trust, what output proves success, and what would go wrong if another layer owned it?
If your explanation is vague, repeat the boundary trace before moving on.Close the reference. Reproduce the core idea from memory or explain the exact request/data flow aloud. Then check your diff or notes for accidental shortcuts that bypass validation, authorization, error handling, or verification earned in earlier phases.