feat(sidebar): display input/cache tokens with cache percentage in context panel

[walker83]

What

Enhance the right-side Context panel to display a breakdown of non-cached input tokens vs cached tokens, along with the cache hit percentage.

Before:

Context
123,456 tokens
45% used
$1.23 spent

After:

Context
123,456 tokens (45%)
Input 100K · Cache 23K
19% cached
$1.23 spent

Why

Cache read pricing is 1/5 to 1/10 of regular input pricing (e.g. Anthropic: $0.30/M vs $3.00/M tokens). Users need immediate visibility into their cache hit ratio to optimize costs.

How

• Modified context.tsx sidebar plugin to extract input, cache (read + write), and cachePercent from the last assistant message tokens
• Merged tokens and context percentage into one line to save space
• Added conditional display for cache breakdown using Locale.number() compact format (1K/1.5M)
• cachePercent is Math.round(cache / totalInput * 100) with denominator guard (totalInput > 0)
• The cache block only renders when input > 0 || cache > 0

Compatibility

This feature works on both MiMo-Code and OpenCode — the change is in the shared TUI sidebar plugin (packages/opencode/src/cli/cmd/tui/feature-plugins/sidebar/context.tsx) and uses only the existing AssistantMessage.tokens data model which is identical across both codebases.