feat: v0.2.0 sprint — ground truth eval, crossover/mutation, checkpointing, similarity guards, dataset loader, CLI commands, extended test coverage

Aggregates all v0.2.0 sprint work (GARAA-30 through GARAA-40) and fixes
2 integration tests that broke when the codebase went async (DSPyLLMAdapter
and full pipeline tests now properly await coroutines).

277 tests pass (260 unit + 17 integration).

Co-Authored-By: Paperclip <noreply@paperclip.ing>

tests/conftest.py

@@ -91,3 +91,27 @@ def mock_proposer_port() -> AsyncMock:
         text="You are a very helpful assistant. Answer the question precisely."
     )
     return port
+
+
+@pytest.fixture
+def mock_crossover_port() -> AsyncMock:
+    """Mock CrossoverPort that combines two parent prompts."""
+    port = AsyncMock()
+
+    async def _crossover(parent_a: Prompt, parent_b: Prompt, task_description: str) -> Prompt:
+        return Prompt(text=f"{parent_a.text} Also, {parent_b.text.lower()}")
+
+    port.crossover = AsyncMock(side_effect=_crossover)
+    return port
+
+
+@pytest.fixture
+def mock_mutation_port() -> AsyncMock:
+    """Mock MutationPort that paraphrases a prompt."""
+    port = AsyncMock()
+
+    async def _mutate(prompt: Prompt, task_description: str, mutation_type: str = "paraphrase") -> Prompt:
+        return Prompt(text=f"[{mutation_type}] {prompt.text}")
+
+    port.mutate = AsyncMock(side_effect=_mutate)
+    return port