feat(ml/serving): inject profile features + sort tasks in tip prompt (#79)

- prompts.py: sort tasks overdue-first → priority desc → age desc before
  rendering into the LLM prompt (same ordering as ml/features/context.py)
- prompts.py: render User profile summary line (completion_rate, dismiss_rate,
  preferred_hour) when profile_features are present
- main.py: add profile_features field to PromptContext; plumb from
  GenerateRequest into the prompt builder via model_copy
- logging_config.py: drop add_logger_name processor (incompatible with
  PrintLoggerFactory — caused test ordering failures)
- test_generate.py: 6 new tests covering sort order, profile rendering,
  partial fields, empty profile, and end-to-end plumbing through /generate

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

ml/serving/tests/test_generate.py

@@ -127,6 +127,46 @@ def test_build_prompt_empty_tasks_no_task_line():
     assert "Generate 2 tips" in prompt
 
 
+def test_build_prompt_tasks_sorted_overdue_first():
+    tasks = [
+        {"content": "Low priority", "priority": 1, "is_overdue": False, "task_age_days": 0},
+        {"content": "Overdue task", "priority": 2, "is_overdue": True, "task_age_days": 3},
+    ]
+    ctx = PromptContext(tasks=tasks, hour_of_day=9)
+    prompt = _build_user_v1(ctx, n=2)
+    assert prompt.index("Overdue task") < prompt.index("Low priority")
+
+
+def test_build_prompt_includes_profile_features():
+    ctx = PromptContext(
+        tasks=[],
+        hour_of_day=14,
+        profile_features={"completion_rate_30d": 0.75, "dismiss_rate_30d": 0.1, "preferred_hour": 9},
+    )
+    prompt = _build_user_v1(ctx, n=1)
+    assert "User profile:" in prompt
+    assert "completion_rate=75%" in prompt
+    assert "dismiss_rate=10%" in prompt
+    assert "preferred_hour=09:00" in prompt
+
+
+def test_build_prompt_no_profile_line_when_empty():
+    ctx = PromptContext(tasks=[], hour_of_day=10, profile_features={})
+    prompt = _build_user_v1(ctx, n=1)
+    assert "User profile:" not in prompt
+
+
+def test_build_prompt_profile_partial_fields():
+    ctx = PromptContext(
+        tasks=[],
+        hour_of_day=10,
+        profile_features={"completion_rate_30d": 0.5},
+    )
+    prompt = _build_user_v1(ctx, n=1)
+    assert "completion_rate=50%" in prompt
+    assert "dismiss_rate" not in prompt
+
+
 @pytest.mark.anyio
 async def test_generate_retry_succeeds_on_second_attempt():
     """First response is invalid JSON; second is valid. Should return 200."""
@@ -271,6 +311,38 @@ async def test_generate_echoes_selected_prompt_version():
     assert resp.json()["prompt_version"] == "v2-mentor"
 
 
+@pytest.mark.anyio
+async def test_generate_passes_profile_features_to_prompt():
+    """profile_features from GenerateRequest should appear in the user message sent to LiteLLM."""
+    fake_items = [{"id": "tip-1", "content": "x", "rationale": "y"}]
+    mock_resp = _litellm_response(fake_items)
+    captured_payload: list[dict] = []
+
+    async def _capture(url, *, json, headers):
+        captured_payload.append(json)
+        return mock_resp
+
+    with patch("main.httpx.AsyncClient") as MockClient:
+        instance = AsyncMock()
+        instance.post = AsyncMock(side_effect=_capture)
+        instance.__aenter__ = AsyncMock(return_value=instance)
+        instance.__aexit__ = AsyncMock(return_value=False)
+        MockClient.return_value = instance
+
+        async with AsyncClient(transport=ASGITransport(app=app), base_url="http://test") as client:
+            resp = await client.post("/generate", json={
+                "user_id": "u1",
+                "n": 1,
+                "profile_features": {"completion_rate_30d": 0.8, "preferred_hour": 10},
+            })
+
+    assert resp.status_code == 200
+    user_msg = captured_payload[0]["messages"][1]["content"]
+    assert "User profile:" in user_msg
+    assert "completion_rate=80%" in user_msg
+    assert "preferred_hour=10:00" in user_msg
+
+
 @pytest.mark.anyio
 async def test_generate_422_on_unknown_prompt_version():
     async with AsyncClient(transport=ASGITransport(app=app), base_url="http://test") as client: