Workspace Boundary — Node.js Regression Study

A signed-in user should only edit a note when their active editor membership matches the note’s stored workspace.

This small Node.js study starts with a deliberately defective test-only handler, runs seven shared cases, then verifies the corrected guard against those same cases. Four failures become passes without breaking the legitimate update. Every denied attempt checks both the response and unchanged stored data.

The evidence board makes the defect, correction and retained behavior inspectable rather than presenting an unexplained green test total.

The same seven cases, before and after the correction.

Independent local synthetic study using trusted injected identity and synchronous in-memory data. This demonstrates authorization reasoning and regression testing; it is not a live authentication service, database or infrastructure audit, or production security assessment. No client or real user data. AI-assisted engineering and presentation.