feat: NanoClaude Phase 2 — EXAONE→Gemma 파이프라인, 큐, 상태 API

- ModelAdapter: 범용 OpenAI-compat 어댑터 (stream/complete/health)
- BackendRegistry: rewriter(EXAONE) + reasoner(Gemma4) 헬스체크 루프
- 2단계 파이프라인: EXAONE rewrite → Gemma reasoning (SSE rewrite 이벤트 노출)
- Fallback: 맥미니 다운 시 EXAONE 단독 모드, stream 중간 실패 시 자동 전환
- Cancel-safe: rewrite 전/후, streaming loop 내, fallback 경로 모두 체크
- Rewrite heartbeat: complete_chat 대기 중 2초 간격 processing 이벤트
- JobQueue: Semaphore(3) 기반 동시성 제한, 정확한 queue position
- GET /chat/{job_id}/status, GET /queue/stats 엔드포인트
- DB: rewrite_model, reasoning_model, rewritten_message 컬럼 추가

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

nanoclaude/services/model_adapter.py

@@ -0,0 +1,109 @@
+"""ModelAdapter — 범용 OpenAI-compat 모델 어댑터."""
+
+from __future__ import annotations
+
+import json
+import logging
+from collections.abc import AsyncGenerator
+
+import httpx
+
+logger = logging.getLogger(__name__)
+
+
+class ModelAdapter:
+    """OpenAI-compatible /v1/chat/completions 백엔드 범용 어댑터.
+    Ollama, MLX 등 모두 동일 인터페이스로 호출."""
+
+    def __init__(
+        self,
+        name: str,
+        base_url: str,
+        model: str,
+        system_prompt: str,
+        temperature: float = 0.7,
+        timeout: float = 120.0,
+    ):
+        self.name = name
+        self.base_url = base_url
+        self.model = model
+        self.system_prompt = system_prompt
+        self.temperature = temperature
+        self.timeout = timeout
+
+    async def stream_chat(self, message: str) -> AsyncGenerator[str, None]:
+        """스트리밍 호출. content chunk를 yield."""
+        payload = {
+            "model": self.model,
+            "messages": [
+                {"role": "system", "content": self.system_prompt},
+                {"role": "user", "content": message},
+            ],
+            "stream": True,
+            "temperature": self.temperature,
+        }
+
+        async with httpx.AsyncClient(timeout=self.timeout) as client:
+            try:
+                async with client.stream(
+                    "POST",
+                    f"{self.base_url}/v1/chat/completions",
+                    json=payload,
+                ) as resp:
+                    if resp.status_code != 200:
+                        body = await resp.aread()
+                        logger.error("%s error %d: %s", self.name, resp.status_code, body.decode())
+                        raise RuntimeError(f"{self.name} 응답 실패 ({resp.status_code})")
+
+                    async for line in resp.aiter_lines():
+                        line = line.strip()
+                        if not line or not line.startswith("data: "):
+                            continue
+                        payload_str = line[len("data: "):]
+                        if payload_str == "[DONE]":
+                            return
+                        try:
+                            chunk = json.loads(payload_str)
+                            delta = chunk.get("choices", [{}])[0].get("delta", {})
+                            content = delta.get("content", "")
+                            if content:
+                                yield content
+                        except (json.JSONDecodeError, IndexError, KeyError):
+                            continue
+
+            except httpx.ConnectError:
+                logger.error("%s connection failed: %s", self.name, self.base_url)
+                raise
+            except httpx.ReadTimeout:
+                logger.error("%s read timeout", self.name)
+                raise
+
+    async def complete_chat(self, message: str) -> str:
+        """비스트리밍 호출. 전체 응답 텍스트 반환."""
+        payload = {
+            "model": self.model,
+            "messages": [
+                {"role": "system", "content": self.system_prompt},
+                {"role": "user", "content": message},
+            ],
+            "stream": False,
+            "temperature": self.temperature,
+        }
+
+        async with httpx.AsyncClient(timeout=self.timeout) as client:
+            resp = await client.post(
+                f"{self.base_url}/v1/chat/completions",
+                json=payload,
+            )
+            resp.raise_for_status()
+            data = resp.json()
+            return data["choices"][0]["message"]["content"]
+
+    async def health_check(self) -> bool:
+        """GET /v1/models — 3초 timeout."""
+        try:
+            async with httpx.AsyncClient(timeout=3.0) as client:
+                resp = await client.get(f"{self.base_url}/v1/models")
+                return resp.status_code < 500
+        except Exception:
+            return False