The AI Tutor will answer grounded in this lesson's cited source (jira_story: GOV-153) and key concepts — authority Informational. Not yet enabled.
Lesson 4 of 4 — RBAC defence-in-depth tests
Cross-Cutting & Hardening · Source: Jira Story GOV-153 (3 pts) · Authority: Informational (the Jira story is authoritative).
Purpose
Verify RBAC holds at both UI and API with negative tests.
Why it matters
You only know RBAC works when the 'should-fail' cases actually fail.
Where this lives
Runs in the Customer AI Governance platform (components: platform). Negative RBAC tests assert UI hiding and API 403 for unauthorized roles. Learn teaches this read-only.
Workflow
- Given an auditor or owner
- When negative tests run
- Then they pass green (UI hidden and API 403)
Worked example
RBAC tests are all 'can-do' assertions. What's missing?
The negative 'must-not' cases — auditor/owner deny paths (UI hidden and API 403) must be tested.
Acceptance criteria (authoritative — GOV-153)
Given an auditor or owner, when negative tests run, then they pass green (UI hidden and API 403).
Common pitfalls
- Testing only happy paths, missing the deny cases.
Knowledge check
- Scenario: RBAC tests are all 'can-do' assertions. What's missing?
- Look for: The negative 'must-not' cases — auditor/owner deny paths (UI hidden and API 403) must be tested.
Key concepts
- RBAC negative tests
- defence-in-depth
- deny-path coverage
For the AI Tutor
- Summary: Verify RBAC holds at both UI and API with negative tests.
- When to use: teaching or answering questions about "rbac defence-in-depth tests".
- When NOT to use: resolving the role itself (that is GOV-103).
- Key concepts: RBAC negative tests, defence-in-depth, deny-path coverage.
Completion criteria
Learner passes the scenario knowledge check and can point to the system state that satisfies GOV-153's acceptance criterion.
Related
- Module: LEARN-MOD-005 (Cross-Cutting & Hardening)
- Provenance: LEARN-MAP-001 · Generation standard: LEARN-STD-002
- Next: (module complete)