workflow-fix: update the two test files' mocked auto-chain fixtures to the

Overview / Motivation

Auto-filed by the watcher urgent-park router (#1681, autonomous_session_watch.urgent_wf_park_pass) from an URGENT (urgency: main-red) parked workflow-fix candidate raised on task #1997. The named test is red on origin/main NOW — every intervening session's Step 9c gate re-classifies it until this fix lands.

Goal

fix tests/test_workload_cmd_env_lint.py::test_launch_auto_flagged_cmd_warns_flags_marker_and_proceeds red on origin/main: update the two test files' mocked auto-chain fixtures to the current lane order — unverified hypothesis, verify at plan time: the #2054 runpod-first promotion and/or #2028 GCP-removal (commit 538e0d7a9b landed on this surface within the window) changed which mock backend the auto chain reaches, so nibi is never consulted in the mocked scenario; fixtures should either make the earlier rungs capacity-miss in the mock or pin an explicit backend.

Workflow gap

• Bug observed (emitter's claim, candidate block): 10 nodes red on PRISTINE main (verified twice by #1997's Step 9c compare pristine-scratch oracle at f2bdd400d430 and a repo-root baseline run: 10 failed / 63 passed): each failing node asserts the auto chain lands on nibi (assert len(nibi.launches) == 1 → 0) in the workload-cmd env/persist lint-arm tests.
• Failing node (router-verified): tests/test_workload_cmd_env_lint.py::test_launch_auto_flagged_cmd_warns_flags_marker_and_proceeds
• Confidence (emitter): medium
• verified-at-filing: uv run pytest tests/test_workload_cmd_env_lint.py::test_launch_auto_flagged_cmd_warns_flags_marker_and_proceeds -q -> rc=1 at main @ f4250f8042 (2026-08-05T13:13:08Z) (FAILED — red confirmed)

Proposed change (candidate diff sketch — refine in planning)

(see the verbatim candidate block under ## Provenance — the router forwards it unmodified and never synthesizes fields)

Scope / surfaces

• Primary target: tests/test_workload_cmd_env_lint.py, tests/test_workload_cmd_persist_lint.py
• Failing node: tests/test_workload_cmd_env_lint.py::test_launch_auto_flagged_cmd_warns_flags_marker_and_proceeds

Constraints / invariants

• Workflow-surface only — never experiment code, configs/, or tasks/.
• This session carries a workflow_fix_target: Provenance line — it MUST NOT auto-route its own subagents' workflow-fix candidates (recursion guard, .claude/rules/workflow-fix-on-bug.md § Recursion guard).

Provenance

• workflow_fix_target: tests/test_workload_cmd_env_lint.py, tests/test_workload_cmd_persist_lint.py
• fingerprint: e48a62527498
• routed-by: autonomous_session_watch urgent-wf-park-router (#1681)

target_file: tests/test_workload_cmd_env_lint.py, tests/test_workload_cmd_persist_lint.py bug_observed: 10 nodes red on PRISTINE main (verified twice by #1997's Step 9c compare pristine-scratch oracle at f2bdd400d430 and a repo-root baseline run: 10 failed / 63 passed): each failing node asserts the auto chain lands on nibi (assert len(nibi.launches) == 1 → 0) in the workload-cmd env/persist lint-arm tests. why_workflow_gap: workflow-invariant-adjacent tests red on trunk force every intervening session's Step 9c gate to re-classify the same pre-existing red (this session burned a 2-file baseline run + a 4-file pristine-oracle compare cycle on it); the Step 10d gate's pass verdicts now rest on workflow-surface pre-existing red. proposed_change: update the two test files' mocked auto-chain fixtures to the current lane order — unverified hypothesis, verify at plan time: the #2054 runpod-first promotion and/or #2028 GCP-removal (commit 538e0d7a9b landed on this surface within the window) changed which mock backend the auto chain reaches, so nibi is never consulted in the mocked scenario; fixtures should either make the earlier rungs capacity-miss in the mock or pin an explicit backend. diff_sketch: | In tests/test_workload_cmd_env_lint.py + tests/test_workload_cmd_persist_lint.py:

• mock chain setup: make runpod (and fellows) rungs capacity-miss so auto still reaches nibi, OR assert on the FIRST consulted backend's launches instead of nibi's;

• re-derive expected lane from router.DEFAULT_AUTO_LANE_ORDER rather than hardcoding nibi as the auto landing rung. confidence: medium related_task: #1997 urgency: main-red failing_test: tests/test_workload_cmd_env_lint.py::test_launch_auto_flagged_cmd_warns_flags_marker_and_proceeds wf_fix: true