workflow-fix: reconcile tests/test_workload_cmd_env_lint.py's mock-backend

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 #1996. 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_kill_switch_env_skips_inline_lint red on origin/main: reconcile tests/test_workload_cmd_env_lint.py's mock-backend launch expectations with the post-#2054 runpod-first auto lane order (or fix the router if the launches are genuinely mis-routed).

Workflow gap

• Bug observed (emitter's claim, candidate block): 7 deterministic tests in tests/test_workload_cmd_env_lint.py fail on the main checkout (and at fetched origin/main base 8f95ad4158) — each _MockBackend nibi expects 1 launch and gets 0 (e.g. test_kill_switch_env_skips_inline_lint, tests/test_workload_cmd_env_lint.py:865).
• Failing node (router-verified): tests/test_workload_cmd_env_lint.py::test_kill_switch_env_skips_inline_lint
• Confidence (emitter): medium
• verified-at-filing: uv run pytest tests/test_workload_cmd_env_lint.py::test_kill_switch_env_skips_inline_lint -q -> rc=1 at main @ 6fe11b459a (2026-08-06T00:03:09Z) (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, src/explore_persona_space/backends/router.py
• Failing node: tests/test_workload_cmd_env_lint.py::test_kill_switch_env_skips_inline_lint

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, src/explore_persona_space/backends/router.py
• fingerprint: b5bbd691dd5b
• routed-by: autonomous_session_watch urgent-wf-park-router (#1681)

target_file: tests/test_workload_cmd_env_lint.py, src/explore_persona_space/backends/router.py bug_observed: 7 deterministic tests in tests/test_workload_cmd_env_lint.py fail on the main checkout (and at fetched origin/main base 8f95ad4158) — each _MockBackend nibi expects 1 launch and gets 0 (e.g. test_kill_switch_env_skips_inline_lint, tests/test_workload_cmd_env_lint.py:865). why_workflow_gap: a workflow-invariant test file is red on origin/main, so every intervening session's Step 9c gate must re-classify the red (the #1643/#1681 fleet-cost class); unverified hypothesis — verify at plan time: the #2054 runpod-first DEFAULT_AUTO_LANE_ORDER reorder (and/or the #1997/#1775 pod-lifecycle change, both recent on this surface) landed without updating these tests' auto-lane launch expectations. proposed_change: reconcile tests/test_workload_cmd_env_lint.py's mock-backend launch expectations with the post-#2054 runpod-first auto lane order (or fix the router if the launches are genuinely mis-routed). diff_sketch: | In the 7 failing tests, either register the mock runpod lane as the expected launch target (post-#2054 order: runpod, fellows, nibi, fir, mila) or pin EPM_AUTO_LANE_ORDER in the fixtures so nibi is first, e.g.:

• monkeypatch.setenv("EPM_AUTO_LANE_ORDER", "nibi,fir,mila") confidence: medium related_task: #1996 urgency: main-red failing_test: tests/test_workload_cmd_env_lint.py::test_kill_switch_env_skips_inline_lint wf_fix: true