feat: Phase 1+2 — project setup and core infrastructure

- requirements.txt, config.yaml, .env, Dockerfile, docker-compose.yml
- app/core: config (YAML+env override), logging (JSON structured),
  exceptions (typed hierarchy), json_utils (Markdown fence stripping)
- app/clients: LLMClient ABC + ZhipuAIClient (run_in_executor),
  StorageClient ABC + RustFSClient (boto3 head_object for size check)
- app/main.py: FastAPI app with health endpoint and router registration
- app/core/dependencies.py: lru_cache singleton factories
- tests/conftest.py: mock_llm, mock_storage, test_app, client fixtures
- pytest.ini: asyncio_mode=auto
- 11 unit tests passing

app/clients/llm/base.py

@@ -0,0 +1,11 @@
+from abc import ABC, abstractmethod
+
+
+class LLMClient(ABC):
+    @abstractmethod
+    async def chat(self, model: str, messages: list[dict]) -> str:
+        """Send a text chat request and return the response content string."""
+
+    @abstractmethod
+    async def chat_vision(self, model: str, messages: list[dict]) -> str:
+        """Send a multimodal (vision) chat request and return the response content string."""

app/clients/llm/zhipuai_client.py

@@ -0,0 +1,37 @@
+import asyncio
+
+from zhipuai import ZhipuAI
+
+from app.clients.llm.base import LLMClient
+from app.core.exceptions import LLMCallError
+from app.core.logging import get_logger
+
+logger = get_logger(__name__)
+
+
+class ZhipuAIClient(LLMClient):
+    def __init__(self, api_key: str) -> None:
+        self._client = ZhipuAI(api_key=api_key)
+
+    async def chat(self, model: str, messages: list[dict]) -> str:
+        return await self._call(model, messages)
+
+    async def chat_vision(self, model: str, messages: list[dict]) -> str:
+        return await self._call(model, messages)
+
+    async def _call(self, model: str, messages: list[dict]) -> str:
+        loop = asyncio.get_event_loop()
+        try:
+            response = await loop.run_in_executor(
+                None,
+                lambda: self._client.chat.completions.create(
+                    model=model,
+                    messages=messages,
+                ),
+            )
+            content = response.choices[0].message.content
+            logger.info("llm_call", extra={"model": model, "response_len": len(content)})
+            return content
+        except Exception as exc:
+            logger.error("llm_call_error", extra={"model": model, "error": str(exc)})
+            raise LLMCallError(f"大模型调用失败: {exc}") from exc