Implement credentialed drill packaging workplan

tests/test_evaluation_scenarios.py

@@ -4,7 +4,7 @@ from pathlib import Path
 
 from phase_memory.adapters import InMemorySemanticIndex
 from phase_memory.contracts import graph_from_markitect
-from phase_memory.evaluation import EVALUATION_REPORT_SCHEMA, evaluation_threshold_report
+from phase_memory.evaluation import EVALUATION_REPORT_SCHEMA, EVALUATION_TREND_SCHEMA, evaluation_threshold_report, evaluation_trend_artifact
 from phase_memory.models import ActivationPlan, MemoryPath
 from phase_memory.retrieval import activation_quality_report, select_event_path
 from phase_memory.runtime import PhaseMemoryRuntime
@@ -102,6 +102,30 @@ def test_evaluation_threshold_report_summarizes_all_scenarios() -> None:
     assert report["diagnostics"] == []
 
 
+def test_evaluation_trend_artifact_tracks_threshold_and_metric_deltas() -> None:
+    data = json.loads((FIXTURES / "evaluation-scenarios.json").read_text(encoding="utf-8"))
+    report = evaluation_threshold_report(data)
+    previous = {
+        "id": "previous",
+        "metrics": {
+            **report["metrics"],
+            "policy_denial_count": report["metrics"]["policy_denial_count"] + 1,
+        },
+    }
+
+    trend = evaluation_trend_artifact(
+        report,
+        previous_report=previous,
+        run_metadata={"run_id": "pytest", "created_at": "2026-05-19T00:00:00+00:00"},
+    )
+
+    assert trend["schema_version"] == EVALUATION_TREND_SCHEMA
+    assert trend["run"]["run_id"] == "pytest"
+    assert trend["threshold_deltas"]["policy_denial_count"] == 0.0
+    assert trend["metric_deltas"]["policy_denial_count"] == -1.0
+    assert trend["diagnostics"][0]["code"] == "evaluation_metric_regressed"
+
+
 def _activation_plan(response):
     data = response["data"]["activation_plan"]
     return ActivationPlan(