feat(backend): add embedding client and update LLM client

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

backend/app/services/embedding_client.py

@@ -0,0 +1,52 @@
+import logging
+from typing import List
+
+import httpx
+
+from app.core.config import Settings
+
+
+logger = logging.getLogger(__name__)
+
+
+class EmbeddingClient:
+    """Async embedding client for OpenRouter-compatible embeddings API."""
+
+    def __init__(self, settings: Settings):
+        self.base_url = settings.embedding_base_url.rstrip("/")
+        self.api_key = settings.embedding_api_key or settings.llm_api_key
+        self.model = settings.embedding_model
+        # Async HTTP client for connection pooling
+        self._client: httpx.AsyncClient | None = httpx.AsyncClient(
+            base_url=self.base_url,
+            timeout=settings.llm_timeout,
+            headers={"Authorization": f"Bearer {self.api_key}"},
+        )
+
+    async def embed(self, texts: List[str]) -> List[List[float]]:
+        if not texts:
+            return []
+        payload = {"model": self.model, "input": texts}
+        try:
+            resp = await self._client.post("/embeddings", json=payload)
+            resp.raise_for_status()
+            data = resp.json()
+            # Common OpenAI-like response shapes
+            if isinstance(data, dict):
+                # OpenRouter/OpenAI style: {"data": [{"embedding": []}, ...]}
+                if "data" in data and isinstance(data["data"], list):
+                    return [item.get("embedding", []) for item in data["data"]]
+                # Alternative: {"embeddings": [[...], [...]]}
+                if "embeddings" in data and isinstance(data["embeddings"], list):
+                    return data["embeddings"]
+            # Fallback: try to extract a flat list
+            if isinstance(data, list):
+                return data  # type: ignore[return-value]
+        except Exception as exc:  # pragma: no cover - network/runtime issues
+            logger.error("Embedding API call failed: %s", exc)
+            raise
+        return []
+
+    async def close(self):
+        if self._client:
+            await self._client.aclose()

backend/app/services/llm_client.py

@@ -1,28 +1,49 @@
+import asyncio
+import logging
+from typing import Optional
+
 import httpx
 
 from app.core.config import Settings
 
 
+class LLMClientError(Exception):
+    pass
+
+
 class LLMClient:
+    """Asynchronous LLM HTTP client with connection pooling."""
+
     def __init__(self, settings: Settings):
+        self.settings = settings
         self.base_url = settings.llm_base_url.rstrip("/")
         self.api_key = settings.llm_api_key
         self.model = settings.llm_model_name
+        self.logger = logging.getLogger(__name__)
+        # Use a single shared AsyncClient for all requests
+        self._client = httpx.AsyncClient(
+            base_url=self.base_url,
+            timeout=settings.llm_timeout,
+            headers={"Authorization": f"Bearer {self.api_key}", "Content-Type": "application/json"},
+        )
 
-    def complete(self, prompt: str, temperature: float = 0.7) -> str:
-        response = httpx.post(
-            f"{self.base_url}/chat/completions",
-            headers={
-                "Authorization": f"Bearer {self.api_key}",
-                "Content-Type": "application/json",
-            },
+    async def complete(self, prompt: str, temperature: float = 0.7) -> str:
+        try:
+            resp = await self._client.post(
+                "/chat/completions",
                 json={
                     "model": self.model,
                     "messages": [{"role": "user", "content": prompt}],
                     "temperature": temperature,
                 },
-            timeout=60.0,
             )
-        response.raise_for_status()
-        data = response.json()
+            resp.raise_for_status()
+            data = resp.json()
             return data["choices"][0]["message"]["content"]
+        except (httpx.TimeoutException, httpx.HTTPStatusError, httpx.RequestError) as exc:
+            self.logger.error("LLM API error: %s", exc)
+            raise LLMClientError from exc
+
+    async def close(self):
+        if self._client:
+            await self._client.aclose()