算力中心新增消耗明细和趋势统计接口

- compute_client.py 新增 user_logs 方法，封装引擎日志接口（支持分页）
- rbac_opc.py 新增两个接口：
  - GET /opc/compute/usage/logs - 消耗明细（分页，统一字段格式为前端期望的 snake_case）
  - GET /opc/compute/usage/trend - 消耗趋势统计（近N天趋势、模型分布、时段分布）
- 趋势接口基于引擎日志数据聚合计算：
  - 近N天消耗趋势（按日期聚合 cost 和 tokens）
  - 模型分布（按模型名聚合调用次数，Top 5 + 其他）
  - 时段分布（按24小时聚合消耗）
- 引擎对接失败时返回空数据，不报错（避免前端白屏）
- 模型分布自动分配颜色，便于前端饼图展示

app/api/routers/rbac_opc.py

@@ -366,6 +366,134 @@ async def opc_compute_usage(user: dict = Depends(require_roles("opc_member"))):
     return data
 
 
+@router.get("/compute/usage/logs", summary="我的消耗明细（分页）")
+async def opc_compute_usage_logs(
+    page: int = 1,
+    page_size: int = 20,
+    user: dict = Depends(require_roles("opc_member")),
+):
+    """消耗流水明细，支持分页。"""
+    try:
+        data = await compute_client.user_logs(user.get("username"), page=page, page_size=page_size)
+    except compute_client.ComputeError as exc:
+        raise HTTPException(status_code=502, detail=f"算力引擎对接失败: {exc}") from exc
+
+    items = data if isinstance(data, list) else data.get("items", [])
+    total = data.get("total", len(items)) if isinstance(data, dict) else len(items)
+
+    # 统一字段格式（snake_case → 前端期望的格式）
+    out = []
+    for it in items or []:
+        it = dict(it)
+        out.append({
+            "id": it.get("id") or it.get("log_id") or "",
+            "time": it.get("created_at") or it.get("time") or it.get("created_time") or "",
+            "model": it.get("model_name") or it.get("model") or "",
+            "type": it.get("type") or it.get("mode") or "chat",
+            "input_tokens": it.get("input_tokens") or it.get("prompt_tokens") or 0,
+            "output_tokens": it.get("output_tokens") or it.get("completion_tokens") or 0,
+            "cost": it.get("cost") or it.get("cost_quota") or 0,
+            "status": it.get("status") or "success",
+        })
+
+    return {"items": out, "total": total, "page": page, "page_size": page_size}
+
+
+@router.get("/compute/usage/trend", summary="消耗趋势统计（近7天/模型分布/时段分布）")
+async def opc_compute_usage_trend(
+    days: int = 7,
+    user: dict = Depends(require_roles("opc_member")),
+):
+    """消耗趋势统计：近N天消耗趋势、模型分布、时段分布。"""
+    from collections import defaultdict
+    from datetime import datetime, timedelta
+
+    try:
+        # 获取最近的消耗日志（最多取 500 条用于统计）
+        data = await compute_client.user_logs(user.get("username"), page=1, page_size=500)
+    except compute_client.ComputeError as exc:
+        # 引擎对接失败时返回空数据，不报错
+        data = {"items": []}
+
+    items = data if isinstance(data, list) else data.get("items", [])
+
+    # 近N天消耗趋势
+    now = datetime.now()
+    trend_map = {}
+    for i in range(days):
+        d = (now - timedelta(days=days - 1 - i)).strftime("%m-%d")
+        trend_map[d] = {"date": d, "cost": 0, "tokens": 0}
+
+    # 模型分布
+    model_dist = defaultdict(lambda: {"name": "", "value": 0, "cost": 0})
+
+    # 时段分布（24小时）
+    hourly_dist = [{"hour": f"{h:02d}", "cost": 0} for h in range(24)]
+
+    total_cost = 0
+    total_tokens = 0
+
+    for it in items or []:
+        it = dict(it)
+        # 解析时间
+        time_str = it.get("created_at") or it.get("time") or it.get("created_time") or ""
+        cost = float(it.get("cost") or it.get("cost_quota") or 0)
+        input_tokens = int(it.get("input_tokens") or it.get("prompt_tokens") or 0)
+        output_tokens = int(it.get("output_tokens") or it.get("completion_tokens") or 0)
+        tokens = input_tokens + output_tokens
+        model_name = it.get("model_name") or it.get("model") or "unknown"
+
+        total_cost += cost
+        total_tokens += tokens
+
+        # 趋势统计
+        if time_str:
+            try:
+                dt = datetime.fromisoformat(time_str.replace("Z", "+00:00")) if "T" in time_str else datetime.strptime(time_str[:10], "%Y-%m-%d")
+                date_key = dt.strftime("%m-%d")
+                if date_key in trend_map:
+                    trend_map[date_key]["cost"] += cost
+                    trend_map[date_key]["tokens"] += tokens
+
+                # 时段统计
+                hour = dt.hour
+                hourly_dist[hour]["cost"] += cost
+            except (ValueError, TypeError):
+                pass
+
+        # 模型分布统计
+        model_dist[model_name]["name"] = model_name
+        model_dist[model_name]["value"] += 1
+        model_dist[model_name]["cost"] += cost
+
+    # 模型分布取 Top 5，其余归为"其他"
+    model_list = sorted(model_dist.values(), key=lambda x: x["value"], reverse=True)
+    if len(model_list) > 5:
+        top5 = model_list[:5]
+        others = model_list[5:]
+        other_value = sum(x["value"] for x in others)
+        other_cost = sum(x["cost"] for x in others)
+        top5.append({"name": "其他", "value": other_value, "cost": other_cost})
+        model_list = top5
+
+    # 模型分布颜色
+    colors = ["#6284ff", "#8f7bff", "#3fb68b", "#e8830c", "#e0526e", "#38a8e0"]
+    for i, m in enumerate(model_list):
+        m["color"] = colors[i % len(colors)]
+
+    return {
+        "trend": list(trend_map.values()),
+        "model_distribution": model_list,
+        "hourly_distribution": hourly_dist,
+        "summary": {
+            "total_cost": total_cost,
+            "total_tokens": total_tokens,
+            "total_requests": len(items or []),
+            "days": days,
+        },
+    }
+
+
 @router.get("/compute/balance", summary="我的算力余额")
 async def opc_compute_balance(user: dict = Depends(require_roles("opc_member"))):
     try:

app/services/compute_client.py

@@ -174,6 +174,19 @@ async def user_usage(username: str) -> dict:
     return data.get("data") or {}
 
 
+async def user_logs(username: str, page: int = 1, page_size: int = 50) -> dict:
+    """用户消耗流水/日志（PAT 鉴权）。"""
+    key = await _user_key(username)
+    if not key:
+        return {"items": [], "total": 0}
+    data = await _request(
+        method="GET",
+        path=f"/api/log/self?page={page}&page_size={page_size}",
+        headers={"Authorization": f"Bearer {key}"},
+    )
+    return data.get("data") or data
+
+
 async def delete_token(token_id: int) -> dict:
     """删除引擎令牌。"""
     return await _request(