test(live-workflow): fix full_live_workflow dodge by using gemini_cli mock

The test was previously marked @pytest.mark.skip because it used
current_provider='gemini' (the real Gemini API). With no API key or
under load, the test aborts with 'AI Status went to error during response
wait'.

Applied the same fix pattern as test_extended_sims.py context_sim_live
et al:
- current_provider: gemini_cli (was: gemini)
- gcli_path: tests/mock_gemini_cli.py (was: not set)
- Removed current_model setting (not needed for the mock)

Verification: tier-3-live_gui PASS in 602s with this test now PASSING
(was: SKIPPED). The test still asserts the full live workflow per the
'ANTI-SIMPLIFICATION' contract in the docstring.

tests/test_live_workflow.py

@@ -29,7 +29,6 @@ def wait_for_value(client, field, expected, timeout=10):
  return False
 
 @pytest.mark.integration
-@pytest.mark.skip(reason="Requires real AI provider (gemini API); 503 UNAVAILABLE on demand spikes aborts the test. Same class of pre-existing flake as test_execution_sim_live (see test_extended_sims.py). Documented in fix_test_failures_20260624 VC4 PARTIAL note.")
 def test_full_live_workflow(live_gui) -> None:
  """
  
@@ -125,10 +124,11 @@ def test_full_live_workflow(live_gui) -> None:
 
  # Enable auto-add so the response ends up in history
  client.set_value("auto_add_history", True)
- client.set_value("current_provider", "gemini")
- 
- # USE gemini-2.0-flash-lite (Actual current model)
- client.set_value("current_model", "gemini-2.5-flash-lite")
+ # Use gemini_cli with the mock script (same pattern as test_extended_sims.py
+ # context/ai_settings/tools sims and as used by the local MockGeminiCli shim)
+ client.set_value("current_provider", "gemini_cli")
+ client.set_value("gcli_path", f'"{sys.executable}" "{os.path.abspath("tests/mock_gemini_cli.py")}"')
+
  time.sleep(1)
 
  # 3. Discussion Turn