test(orchestrator): Align post-merge CTM fixtures

[mwolson]

Summary

• Align explicit server-home assertions with the orchestrator v2 userdata-v2
  state directory.
• Update two Claude replay query.open frames for the current read-only tool
  availability policy.
• Restore the CTM Test job after the [orchestrator-v2] fix(orchestrator): Harden Grok v2 settlement, background task lifecycle, steer visibility, and images #3578 merge and related Claude policy
  updates.

Problem and Fix

Problem and Why it Happened	Fix
Two CLI configuration tests still expected the pre-v2 userdata directory even though explicit server paths resolve under userdata-v2.	Update the state directory and database path assertions to match production path derivation.
Two Claude replay transcripts still expected the unrestricted claude_code tool preset after CTM began passing explicit read-only tool lists.	Update only the recorded query.open frames to expect Read, Glob, and Grep, including allowedTools where the runtime policy supplies it.

Validation

• Latest-base focused CLI config, desktop environment, and orchestrator replay
  fixtures: 17 tests pass, 1 skipped
• vp check: pass with existing warnings only
• vp run typecheck: pass
• Grok implementation self-review: pass
• Codex independent review: no findings

Note

Align CTM test fixtures to use explicit tool lists and userdata-v2 state directory

• Updates claude_transcript.ndjson fixtures in tool_call_read_only_on_request and tool_call_restricted_granular to expect an explicit ["Read", "Glob", "Grep"] tool list instead of a {type: "preset", preset: "claude_code"} object; tool_call_read_only_on_request also adds allowedTools to the expected payload.
• Updates config.test.ts assertions to expect stateDir and dbPath under userdata-v2 instead of userdata.

^{Macroscope summarized 462e3b4.}

---

Note

Low Risk
Test and replay fixture updates only; no runtime logic, auth, or data migration in this diff.

Overview
Brings CTM / post-merge tests back in line with current server path derivation and Claude orchestrator tool wiring—no production behavior changes.

CLI config tests now assert stateDir and dbPath under userdata-v2 instead of userdata, matching deriveServerPaths when the base dir is explicit.

Claude replay fixtures (tool_call_read_only_on_request, tool_call_restricted_granular) update the expected query.open frame: replace the claude_code tool preset with explicit ["Read","Glob","Grep"], and add allowedTools on the read-only-on-request scenario where the runtime supplies it.

^{Reviewed by Cursor Bugbot for commit 462e3b4. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: e5f4192d-63cc-431c-8af3-713b12f4e8d4

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[macroscopeapp]

Approvability

Verdict: Approved

All changes are confined to test files and test fixtures, with no production code modified. The updates align expected values after a merge, with zero runtime behavior impact.

^{You can customize Macroscope's approvability policy. Learn more.}