Add Force Stop Capability to Interrupt Agent Reasoning or Tool Calls #2853

** Please make sure you read the contribution guide and file the issues in the right place. **
Contribution guide.

Is your feature request related to a problem? Please describe.
Currently, once an agent starts thinking, reasoning, or calling a tool, there is no way to stop it mid-execution. This can lead to wasted time and resources if the reasoning loop takes too long or when the model hallucinates, if a tool call hangs, or if the user realizes they want to cancel the current request.

Describe the solution you'd like
Introduce a "force stop" mechanism that allows users (or the calling application) to immediately halt an agent’s reasoning process or tool invocation. This could be an API call, a UI button (for interactive sessions), or a timeout mechanism that can safely interrupt ongoing reasoning/tool usage without corrupting state.

Describe alternatives you've considered

• Relying on timeouts at the system level (not ideal, as it doesn’t provide fine-grained control).
• Restarting the entire agent session
• Manually killing the process

Additional context
A force stop option would greatly improve developer and user experience, especially in long-running workflows, debugging scenarios, or cases where an agent mistakenly enters repetitive or non-terminating reasoning.