Phase 0.7 — Acceptance Gate 자동화 + LM Studio 통합 + OntoCast 버그 수정

- platform/ → ont_platform/ rename
  Python 내장 platform 모듈과 이름 충돌. numpy/scipy가 platform.machine() 호출 시
  우리 패키지를 가져와 AttributeError. ont_platform으로 변경하고 pyproject.toml,
  ont_platform/**, tests/** import 경로 모두 업데이트.

- ont_platform/config.py: lenient LLM builder 추가
  LM Studio/vLLM 등 OpenAI-호환 로컬 서버가 임의 모델 식별자(예: deepseek-r1-distill-
  qwen-7b)를 쓸 수 있도록 OntoCast의 OpenAIModel enum validation을 Pydantic
  model_construct로 우회. ToolConfig() 생성 시 충돌을 막기 위해 LLM_MODEL_NAME을
  잠시 비웠다가 lenient 인스턴스로 교체.

- ont_platform/api/deps.py: ToolBox 초기화를 asyncio.to_thread로 격리
  LLMTool.create()가 내부에서 asyncio.run()을 부르는데 lifespan/테스트가 이미
  async 컨텍스트라 이중 loop 충돌. 별도 스레드에서 sync 생성자 실행.

- 테스트 인프라 정비
  * tests/integration/test_api_smoke.py: TestClient 구버전 starlette 호환을 위해
    lifespan='off' 대신 app.router.lifespan_context = noop 패턴 적용.
  * tests/unit/test_convert_document.py, test_select_ontology.py: ontocast.agent
    __init__.py가 re-export한 함수가 서브모듈을 가리는 문제로 sys.modules에서
    실제 모듈 객체 직접 추출.
  * tests/e2e/conftest.py: .env 자동 로드 + provider별 skip 조건 (Ollama는
    LLM_API_KEY 불필요).
  * tests/e2e/test_phase0_full_pipeline.py: provider별 키 분기,
    HDBSCAN 클러스터링이 동작하도록 fixture 페이로드 16문장으로 확장.

- vendored OntoCast 버그 수정 3건 (VENDORED_MODIFICATIONS.md 기록):
  * agent/render_ontology.py: render_ontology_fresh()의 .format() 호출에 누락된
    ontology_prefix 인자 추가 (Bootstrap 단계에서 KeyError: 'ontology_prefix').
  * stategraph/node_factories.py: render_ontology/render_facts 노드의
    state.model_copy(deep=True)로 budget_tracker가 deep-copy되어 root state의
    BudgetTracker가 영원히 0인 채로 남던 버그 수정. 원본 인스턴스 공유로 변경.

- 문서 갱신
  README.md (Phase 0.7 부분완료 + ont_platform 폴더 이름),
  docs/phases/PHASE0_ACCEPTANCE_GATE.md (검증 이력 + Ollama/LM Studio 옵션),
  .env.example (LM Studio/Ollama/OpenAI 세 옵션 명시).

검증
- unit + integration 26/26 통과.
- e2e (LM Studio + Qwen3-8B / DeepSeek-R1-Distill-Qwen-7B): 워크플로우 끝까지
  실행 + 5번 LLM 호출 + LangGraph 전 노드 traceable 확인. 7-8B 로컬 모델은
  strict structured output(Turtle RDF in JSON) 한계로 ontology/facts TTL 자동
  생성 부분 성공. 클라우드 LLM 환경에서 재검증 필요.

Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>

ontology_platform/ont_platform/api/deps.py

@@ -0,0 +1,119 @@
+"""FastAPI dependencies and app lifespan.
+
+Owns the ToolBox singleton, the compiled LangGraph workflow, and the
+recursion limit. Construction happens once at app startup; teardown is a
+no-op because OntoCast tools don't currently expose a close hook.
+
+Why a module-level holder instead of `app.state`:
+    The /process route is async and may be entered concurrently. A simple
+    dict on `app.state` works, but pulling the ToolBox through a typed
+    dependency function makes the contract explicit and gives every route
+    the same view of "what tools and what workflow are available right now".
+"""
+
+from __future__ import annotations
+
+import asyncio
+import logging
+import sys
+from dataclasses import dataclass
+from pathlib import Path
+
+# Make vendored OntoCast importable. (Same trick as platform/config.py.)
+_REPO_ROOT = Path(__file__).resolve().parents[2]
+_VENDORED_ONTOCAST = _REPO_ROOT / "vendored" / "ontocast"
+if str(_VENDORED_ONTOCAST) not in sys.path:
+    sys.path.insert(0, str(_VENDORED_ONTOCAST))
+
+from langchain_core.runnables import RunnableConfig  # noqa: E402
+from langgraph.graph.state import CompiledStateGraph  # noqa: E402
+
+from ontocast.cli.serve import calculate_recursion_limit  # noqa: E402
+from ontocast.config import ServerConfig  # noqa: E402
+from ontocast.stategraph import create_agent_graph  # noqa: E402
+from ontocast.toolbox import ToolBox  # noqa: E402
+
+import importlib
+
+platform_config = importlib.import_module("ont_platform.config")
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class AppContext:
+    """Immutable holder for objects that live as long as the FastAPI app."""
+
+    settings: "platform_config.PlatformSettings"
+    tools: ToolBox
+    workflow: CompiledStateGraph
+    server_config: ServerConfig
+    recursion_limit: int
+
+
+_context: AppContext | None = None
+
+
+async def initialize_app_context(
+    settings: "platform_config.PlatformSettings | None" = None,
+    *,
+    head_chunks: int | None = None,
+) -> AppContext:
+    """Build ToolBox + workflow once. Idempotent on repeat calls."""
+    global _context
+    if _context is not None:
+        return _context
+
+    settings = settings or platform_config.load_settings()
+    ontocast_config = platform_config.build_ontocast_config(settings)
+
+    # ToolBox.__init__ 내부에서 LLMTool.create()가 `asyncio.run()`을 호출한다.
+    # FastAPI lifespan/테스트가 이미 async 컨텍스트면 이중 loop 충돌이 나므로,
+    # 별도 스레드에서 sync 생성자를 실행한다.
+    tools = await asyncio.to_thread(ToolBox, ontocast_config)
+    # OntoCast's ToolBox.initialize is async; do it here so a request doesn't
+    # have to pay the cost.
+    await tools.initialize()
+
+    workflow = create_agent_graph(tools)
+    server_config = ontocast_config.server
+    recursion_limit = calculate_recursion_limit(head_chunks, server_config)
+
+    _context = AppContext(
+        settings=settings,
+        tools=tools,
+        workflow=workflow,
+        server_config=server_config,
+        recursion_limit=recursion_limit,
+    )
+    logger.info(
+        "App context initialized "
+        f"(phase={int(settings.phase)}, backend={settings.storage_backend}, "
+        f"working_dir={settings.working_directory})"
+    )
+    return _context
+
+
+def get_app_context() -> AppContext:
+    """FastAPI dependency. Raises clearly if the lifespan never ran."""
+    if _context is None:
+        raise RuntimeError(
+            "AppContext not initialized — the app lifespan must run "
+            "`initialize_app_context()` before serving requests."
+        )
+    return _context
+
+
+def reset_app_context_for_testing() -> None:
+    """Test-only: drop the cached context so each test can rebuild it."""
+    global _context
+    _context = None
+
+
+__all__ = [
+    "AppContext",
+    "RunnableConfig",
+    "get_app_context",
+    "initialize_app_context",
+    "reset_app_context_for_testing",
+]