[Bug] Python SDK: ModelClient polling loops lack timeout and cancellation support

**Bug Description**

`ModelClient` 中的轮询循环 (`while True`) 缺少超时机制和取消支持，可能导致程序永久阻塞。

在 PR #492 (TypeScript SDK) 的 Code Review 中，Reviewer @dengwx2026 指出：
> **Add cancellation support**: `ModelClient` polling loops (`while(true)`) need `AbortController` or timeout mechanisms to avoid blocking forever.

经调研，该问题源自 Python SDK 的原始设计。

**Steps to Reproduce**

1. 创建 `ModelClient` 实例
2. 调用 `anti_call_llm(index=0)` 方法
3. 如果日志文件不存在或永远没有请求写入，程序将无限阻塞

**Expected Behavior**

- 应该支持超时参数，允许设置最大等待时间
- 应该支持取消操作（如 `asyncio.CancelledError`）
- 在超时或取消时抛出明确的异常

**Actual Behavior**

- 程序在 `while True` 循环中无限等待
- 无超时机制
- 无取消支持
- 无法中断阻塞状态

**Error Logs**

无错误日志输出，程序静默阻塞。

**Environment Information**

- **OS**: macOS 14.0 (Darwin 23.6.0)
- **Python Version**: 3.14.2
- **ROCK Version**: dev (feature/ts-sdk branch)
- **Installation Method**: source installation
- **Docker Version**: N/A
- **Deployment Type**: local

**ROCK Configuration**

- **Runtime Environment Type**: N/A
- **Sandbox Image**: N/A
- **Resource Allocation**: N/A

**Component Affected**

- [ ] Sandbox
- [ ] Actions
- [ ] Deployments
- [x] SDK & API
- [ ] Envhub
- [ ] CLI
- [ ] Performance & Optimization
- [ ] Documentation & Examples

---

**Additional Context**

问题代码位于 `sdk/model/client.py`：

1. `pop_request` 方法 (第67-76行)
2. `wait_for_first_request` 方法 (第107-120行)

**Suggested Solutions**

```python
# 方案一：添加超时参数
async def pop_request(self, index: int, timeout: float | None = None) -> str:
start_time = time.monotonic()
while True:
if timeout is not None and time.monotonic() - start_time > timeout:
raise TimeoutError(f"pop_request timed out after {timeout} seconds")
# ... rest of the logic
```

```python
# 方案二：支持 asyncio.CancelledError
async def pop_request(self, index: int) -> str:
while True:
try:
# ... existing logic
await asyncio.sleep(1)
except asyncio.CancelledError:
logger.info("pop_request cancelled")
raise
```

```python
# 方案三：调用方使用 asyncio.wait_for 包装
request = await asyncio.wait_for(client.pop_request(index), timeout=30.0)
```

Start in sdk/model/client.py with pop_request (lines 67–76) and wait_for_first_request (lines 107–120), then trace how anti_call_llm(index=0) reaches those polling loops. Clarify the timeout and cancellation behavior against the three suggested approaches. Done means the loops no longer block forever and timeout or cancellation produces a clear exception.

Written by the indexing model from the issue text.