Feature Request: Real-time backend health/trust indicator in CLI

Problem

When Anthropic's backend is overloaded or degraded, Claude Code users have no way to know this. The result:

• Users waste significant time trying to get useful work done
• Sessions produce poor quality output without any indication why
• The external status page (status.anthropic.com) often shows "operational" even during degradation
• Users lose trust in the tool because they can't distinguish "Claude having a bad day" from "my prompt is wrong"

Previous related issues (#5815, #1838, #3628) were closed as NOT_PLANNED, but this remains a critical gap for professional use.

Proposed Solution

Add a visible health/trust indicator to the Claude Code CLI that shows:

1. Backend status: Connection quality, 529 error rate, response latency
2. Session quality: Context usage, degradation warnings
3. Proactive alerts: "Backend is slow right now - consider pausing work"

This could integrate with the existing statusline feature.

Why This Matters

Users paying for Claude Code need to trust it for productive work. Without visibility into backend health, users cannot make informed decisions about whether to:

• Continue working
• Wait for backend to recover
• Switch to alternative tools

A simple traffic-light indicator (green/yellow/red) would be enough to save users hours of wasted effort.