[BUG] Allow configuring [goal] check timeout to prevent hangs on slow network/throughput

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

When running Claude Code under a low-throughput, high-latency, or slow network environment, the independent Evaluator process for /goal takes an excessively long time to analyze the terminal state.
Eventually, it hits an undocumented, hardcoded timeout and pauses the task with the following message:

Goal paused · the goal check timed out · send a message to continue

Since this timeout limit is not documented and cannot be configured, it forces users on slower connections or larger codebases into frequent, frustrating pauses without any way to adjust the threshold.

What Should Happen?

1. Claude Code should provide a configuration parameter (e.g., via claude config or a CLI flag like --goal-check-timeout) to allow users to increase or customize the goal check timeout duration.
2. The default timeout value and behavior should be clearly documented.

Error Messages/Logs

Goal paused · the goal check timed out · send a message to continue

Steps to Reproduce

1. Use a model with heavy latency/low throughput.
2. Initialize an automatic task using the /goal command (e.g., /goal refactor file.js and run tests).
3. Wait for the task to execute steps. When the model invokes the background evaluator to check the current progress, the process will hang for several minutes.
4. The task will abruptly pause with the goal check timed out due to the slow API throughput.

Claude Model

Not sure / Multiple models

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.1.281.0

Platform

Anthropic API

Operating System

Windows

Terminal/Shell

Windows Terminal

Additional Information

_No response_