feat(skills): install deepeval skill from confident-ai/deepeval

Adds the deepeval eval-loop skill via `npx skills add` with skills-lock.json
for reproducible reinstalls. Symlinked to Claude Code via .claude/skills/.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016z2ZFYHQCex8yZAMVMTZzZ

.agents/skills/deepeval/templates/test_single_turn_no_tracing.py

@@ -0,0 +1,32 @@
+from importlib import import_module
+
+import pytest
+
+from deepeval import assert_test
+from deepeval.dataset import EvaluationDataset, Golden
+from deepeval.test_case import LLMTestCase
+
+from metrics import SINGLE_TURN_NO_TRACING_METRICS
+
+
+ai_app = import_module("ai_app")
+
+
+dataset = EvaluationDataset()
+dataset.add_goldens_from_json_file(file_path="tests/evals/.dataset.json")
+
+
+@pytest.mark.parametrize("golden", dataset.goldens)
+def test_single_turn_no_tracing(golden: Golden):
+    actual_output = ai_app.run_ai_app(golden.input)
+    test_case = LLMTestCase(
+        input=golden.input,
+        actual_output=actual_output,
+        expected_output=getattr(golden, "expected_output", None),
+        context=getattr(golden, "context", None),
+        retrieval_context=getattr(golden, "retrieval_context", None),
+    )
+    assert_test(
+        test_case=test_case,
+        metrics=SINGLE_TURN_NO_TRACING_METRICS,
+    )