refactor: 修复候选分页逻辑并拆分配额检查模块

- list_all_candidates 返回 provider_batch_count，区分"无候选"与"无 Provider"
  避免分页在有 Provider 但无候选时提前终止
- 将配额检查逻辑从 aware_scheduler.py 拆分到 quota_skipper.py
- 提取 reorder_candidates 方法，支持跨页汇总后全局重排序
- CandidateResolver 分页循环改用 provider_batch_count 判断终止
- candidate extra_data 中新增 mapping_matched_model 字段
- 新增契约测试和分页行为测试

tests/contracts/test_public_api_routes_contract.py

@@ -0,0 +1,40 @@
+from __future__ import annotations
+
+from fastapi import FastAPI
+
+
+def _build_contract_app() -> FastAPI:
+    from src.api.public.claude import router as claude_router
+    from src.api.public.gemini import router as gemini_router
+    from src.api.public.openai import router as openai_router
+
+    app = FastAPI()
+    app.include_router(claude_router)
+    app.include_router(openai_router)
+    app.include_router(gemini_router)
+    return app
+
+
+def test_public_api_routes_contract_paths_and_tags() -> None:
+    app = _build_contract_app()
+    schema = app.openapi()
+
+    paths = schema.get("paths") or {}
+
+    expected = [
+        ("/v1/messages", "post", "Claude API"),
+        ("/v1/messages/count_tokens", "post", "Claude API"),
+        ("/v1/chat/completions", "post", "OpenAI API"),
+        ("/v1/responses", "post", "OpenAI API"),
+        ("/v1beta/models/{model}:generateContent", "post", "Gemini API"),
+        ("/v1beta/models/{model}:streamGenerateContent", "post", "Gemini API"),
+        ("/v1/models/{model}:generateContent", "post", "Gemini API"),
+        ("/v1/models/{model}:streamGenerateContent", "post", "Gemini API"),
+    ]
+
+    for path, method, expected_tag in expected:
+        assert path in paths, f"missing path {path}"
+        operations = paths.get(path) or {}
+        assert method in operations, f"missing {method.upper()} {path}"
+        tags = operations.get(method, {}).get("tags") or []
+        assert expected_tag in tags, f"{method.upper()} {path} missing tag {expected_tag}"