[Improvement] Add agent activity monitoring and status updates to improve UX

Issue Type: Feature Request

Summary:

Claude Code agents currently run in a "black box" with no visibility into their progress, status, or activity. This creates a poor user experience where users don't know if agents are working, stuck, or have failed silently.

Problem

When running agents (e.g., /review for code review), users experience:

• No progress indicators or status updates
• No way to know if the agent is still working or has crashed
• No time estimates or elapsed time display
• No option to cancel/restart stuck agents
• Uncertainty about whether to wait or try again

Suggested Solution

Add agent activity monitoring with:

Progress Indicators
review is reviewing pull request…
├── ✓ Fetching PR details
├── ⏳ Analyzing diff (2.1MB)
├── ⏳ Checking code quality
└── ⏸ Generating review

Status Features

• Current step being executed
• Elapsed time counter
• Progress percentage when possible
• Cancel option (Ctrl+C or similar)
• Error messages if something fails
• Queue status if multiple agents are running

Minimal Implementation
Even simple text updates would help:
review is reviewing pull request… (Fetching details)
review is reviewing pull request… (Analyzing 46 files)
review is reviewing pull request… (Generating review)

### Impact
This would significantly improve the Claude Code experience by:

• Reducing user anxiety about agent status
• Preventing unnecessary restart attempts
• Providing feedback for long-running operations
• Improving overall UX and perceived performance

Environment

• Claude Code version: [Current]
• OS: Windows/macOS/Linux
• Use case: Code review agent, but applies to all agents