test(surreal_container): additive seam falsifier — SurrealMailboxView → VersionScheduler → KanbanMove

[AdaWorldAPI]

What

Adds a seam falsifier for the SurrealMailboxView → VersionScheduler::on_version → KanbanMove path — the IN-direction of the surreal↔kanban↔scheduler wiring.

Additive: one new test file, zero source change. It is also the first integration-test file in surreal_container (the crate had none).

Why

The scheduler→kanban→view path (a Lance version tick lowering to the next Rubicon KanbanMove over a MailboxSoaView) existed only as doc-comment prose plus one single-tick in-crate unit test against a hand-rolled FakeView. Nothing asserted the real SurrealMailboxView read-glove across the full Rubicon lifecycle. This puts that seam on a test bench.

Each test is kill-condition-first — it pins the contract the doc-comment claims, not current behaviour (deliberately avoiding the stale-test trap that produced the 3 stale bgz-tensor assertions fixed in #506):

Test	Kill-condition (what would prove the seam WRONG)
full_rubicon_arc_lowers_to_legal_successors	a forward tick emits an edge the Rubicon DAG doesn't sanction (can_transition_to == false); walks Planning→CognitiveWork→Evaluation→Commit + Plan→Planning
absorbing_columns_schedule_no_move	Commit/Prune schedule an advance → a non-terminating lifecycle
libet_anchor_only_on_sigma_commit_crossing	the -550_000µs readiness anchor lands anywhere but the Planning→CognitiveWork Σ-commit crossing
lowering_is_deterministic	same (view, version, exec) → different move → hidden state leaked in
exec_target_rides_onto_the_move	the ExecTarget backend selector is dropped/rewritten during lowering

Scope (deliberate)

IN-direction seam only — contract-space (SurrealMailboxView + NextPhaseScheduler, both reachable without the gated surrealdb cold-build path). This does not touch the contested planner-emit half (the CognitiveCycle sequencer + the §9 LOCKED decisions from #496), which is the planner-owning session's lane. This bench is the template that half can be verified against once it lands.

Verification (local — sandbox has the full lance-7 toolchain)

cargo test -p surreal_container --test scheduler_seam   → 5 passed / 0 failed
cargo fmt  -p surreal_container -- --check              → clean

Note

surreal_container is not in any CI workflow, so — like the crate itself — this test runs locally / on-demand only. Another data point for the standing CI-coverage-gap follow-up (CI tests 4 of ~30 crates; two CI-invisible reds — causal-edge, bgz-tensor — have already been found and fixed this session).

---

Generated by Claude Code

Summary by CodeRabbit

• Tests
  • Added comprehensive test coverage for scheduler functionality to verify expected behavior and ensure system reliability.

[coderabbitai]

📝 Walkthrough

Walkthrough

A new test file crates/surreal_container/tests/scheduler_seam.rs is added, containing a view_at helper and five unit tests that verify the NextPhaseScheduler::on_version lowering pipeline: transition arc legality, absorbing-column quiescence, Libet offset placement, determinism, and ExecTarget passthrough.

Changes

Scheduler Seam Falsifier Tests

Layer / File(s)	Summary
Test scaffold and transition legality crates/surreal_container/tests/scheduler_seam.rs	Adds module-level docs, a view_at(phase) helper constructing a minimal SurrealMailboxView, full_rubicon_arc_lowers_to_legal_successors asserting each emitted move is sanctioned by can_transition_to, and absorbing_columns_schedule_no_move asserting on_version returns None for Commit/Prune phases.
Behavioral property tests crates/surreal_container/tests/scheduler_seam.rs	Adds libet_anchor_only_on_sigma_commit_crossing verifying libet_offset_us is -550_000 only on the Planning→CognitiveWork crossing and 0 for all other non-absorbing sources; lowering_is_deterministic asserting identical inputs produce equal moves; and exec_target_rides_onto_the_move asserting the ExecTarget selector is preserved unchanged in the emitted move.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

• AdaWorldAPI/lance-graph#439: Introduced the can_transition_to / absorbing-column contract and KanbanMove.exec: ExecTarget tagging that these seam tests directly exercise and assert against.

Poem

🐇 A rabbit hops through columns five,
From Planning's dawn to where tasks thrive,
No move from Commit — absorbing stays,
Libet fires once through σ-commit haze,
The exec tag rides safe, never dropped,
Determinism locked, no flip-flopped!

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly summarizes the main change: adding a seam test suite for the SurrealMailboxView → VersionScheduler → KanbanMove integration, matching the file addition and PR objectives.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

---

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 and usage tips.}

[coderabbitai]

🧹 Nitpick comments (1)

crates/surreal_container/tests/scheduler_seam.rs (1)

37-49: ⚡ Quick win

Consider verifying mailbox and witness_chain_position passthrough.

The test verifies that mv.from, mv.to, and the legality of the transition are correct, but does not check that mv.mailbox matches view.mailbox_id() or that mv.witness_chain_position matches view.current_cycle(). According to the scheduler implementation, these fields should be populated from the view.

Since the test suite includes a dedicated test for exec passthrough (lines 117-125), consider adding similar verification for mailbox and witness_chain_position to ensure these fields are correctly wired through the lowering pipeline.

Suggested assertion additions

         assert_eq!(mv.from, *from, "move.from must echo the observed phase");
         assert_eq!(mv.to, *want_to, "{from:?} must lower to {want_to:?}");
+        assert_eq!(mv.mailbox, 7, "move.mailbox must match view.mailbox_id()");
+        assert_eq!(mv.witness_chain_position, 13, "move.witness_chain_position must match view.current_cycle()");
         // The DAG itself must sanction the emitted edge — the falsifier.

Alternatively, extract a dedicated test similar to exec_target_rides_onto_the_move that verifies mailbox and witness_chain_position passthrough across multiple phases.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@crates/surreal_container/tests/scheduler_seam.rs` around lines 37 - 49, The
test loop iterating through arc.iter() verifies that mv.from, mv.to, and
transition legality are correct, but omits assertions for the mailbox and
witness_chain_position fields of the Move returned by on_version. Add assertions
within the loop to verify that mv.mailbox matches view_at(*from).mailbox_id()
and that mv.witness_chain_position matches view_at(*from).current_cycle(),
ensuring these fields are correctly propagated from the view through the
scheduler's lowering pipeline alongside the existing transition assertions.

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Nitpick comments:
In `@crates/surreal_container/tests/scheduler_seam.rs`:
- Around line 37-49: The test loop iterating through arc.iter() verifies that
mv.from, mv.to, and transition legality are correct, but omits assertions for
the mailbox and witness_chain_position fields of the Move returned by
on_version. Add assertions within the loop to verify that mv.mailbox matches
view_at(*from).mailbox_id() and that mv.witness_chain_position matches
view_at(*from).current_cycle(), ensuring these fields are correctly propagated
from the view through the scheduler's lowering pipeline alongside the existing
transition assertions.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 70a8e946-4858-46b5-98cc-294053cf1da8

📥 Commits

Reviewing files that changed from the base of the PR and between f7734ff and 010020d.

📒 Files selected for processing (1)

• crates/surreal_container/tests/scheduler_seam.rs