test(ai-lighting): finish Scene→Look rename in test fixtures

[bbernstein]

Summary

• Bring the AI lighting service tests in line with the Scene→Look rename: rename Look-name mock data and corresponding describe/test labels.
• Fix the stale sceneId field in the generateCueSequence mock to lookId, matching the CueData type.
• Preserve theatrical-context strings ('Act 1, Scene 1', lookDescription: 'Romantic scene') per the rename plan's Phase 6.4 — those refer to script scenes, not lighting state.

Test plan

• npm test (all 653 tests pass)
• npm run lint

🤖 Generated with Claude Code

[claude]

Code Review

Overview

This is a focused, test-only PR that completes the Scene→Look rename in tests/services/ai-lighting.test.ts. It has 26 additions and 26 deletions — a pure rename with one meaningful correctness fix.

---

What's Good

• Focused scope: Every change is intentional and part of a coherent rename pass. No unrelated noise.
• Correctness fix for sceneId → lookId: This is the most substantive change and it matters. The generateCueSequence mock had a stale sceneId: '0' field that didn't match the CueData type. Fixing it here prevents a subtle false-positive in the test.
• Theatrical context strings correctly preserved: Keeping 'Act 1, Scene 1' (script scene, not lighting state) and similarly-scoped strings shows good judgment — it distinguishes between the lighting domain rename and the theatrical script language.
• Variable renames (scene/scenes → look/looks): Makes local variable names consistent with the domain terminology, which will reduce reader confusion.
• All 653 tests pass + lint clean: Nothing regresses.

---

Suggestions / Minor Notes

• description: 'A romantic lighting scene' → 'A romantic lighting look': This was changed in the mock data, which is correct — it's describing a lighting state, not a theatrical scene. Just confirming this was intentional (it is consistent with the rename goal).

• Consider a follow-up grep for any remaining sceneId stragglers in other test files: The sceneId fix here was in one mock, but if CueData uses lookId throughout the production type, it's worth a quick scan to confirm no other test fixtures carry stale sceneId references. Not a blocker for this PR, but worth a note.

  grep -r "sceneId" tests/

• No test behavior changed: This is worth stating explicitly — the PR doesn't add, remove, or alter any assertions. All changes are to mock data values and describe/it label strings. Risk is very low.

---

Summary

Clean, low-risk PR. The sceneId → lookId fix is the only real correctness change and it's correct. Everything else is cosmetic alignment. Approve with confidence.

[Copilot]

Pull request overview

Updates the AI lighting service test fixtures to fully reflect the ongoing Scene → Look rename, ensuring test data matches current types and naming.

Changes:

• Renamed test labels and mock look names/descriptions from “Scene” to “Look” for generateLook and additive-look coverage.
• Updated the generateCueSequence mock cue payload to use lookId (instead of stale sceneId) to match the expected cue shape.
• Renamed the scenes fixture array to looks in the generateCueSequence test while preserving theatrical script-context strings (e.g., Act 1, Scene 1).

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[bbernstein]

All review comments have been addressed:

• Copilot's review left no inline comments — only a summary, which described the changes accurately.
• Claude review's optional follow-up (grep for stale sceneId references in other test files) was performed: no other occurrences found in src/ or tests/.

All checks are passing (Lint, Test, Type Check, Test Coverage, Build, Docker Build, Security Audit, MCP Server Validation, claude-review). mergeStateStatus is CLEAN and the PR is ready for merge.