ClaudeAgentOptions.effort typings are outdated, and CLI validation failures are surfaced as opaque ProcessError

## Summary

`ClaudeAgentOptions.effort` in the Python SDK does not include `"xhigh"` in its type definition, even though current Claude Code CLI versions support it.

In addition, when the SDK is used with a Claude Code CLI version that does not support a newer `--effort` value, the failure is surfaced only as a generic `ProcessError` with:

> Command failed with exit code 1
> Error output: Check stderr output for details

This makes debugging unnecessarily difficult.

## Why this matters

There are currently two separate issues:

1. **Outdated Python type definitions**
- `ClaudeAgentOptions.effort` is typed as:
- `Literal["low", "medium", "high", "max"] | None`
- but current Claude Code CLI help includes:
- `low, medium, high, xhigh, max`

2. **Poor error surfacing**
- the SDK forwards `--effort ` directly to the CLI
- if the installed CLI is older and rejects that value, the SDK raises only a generic `ProcessError`
- the actual stderr reason is not included in the exception text

This creates a confusing failure mode:
- newer effort values appear unsupported from the Python types
- and if users try them anyway, failures are hard to diagnose

## Affected version

Observed with:

- `claude-agent-sdk==0.1.61`

## Relevant code locations

### Outdated type definition

In `claude_agent_sdk/types.py`:

```python
effort: Literal["low", "medium", "high", "max"] | None = None