[Feature]: AI token usage meter / daily quota display

[MRAKS05]

Problem

The app tracks daily Gemini API usage (SettingsManager.kt, lines 105-137) with a 20-request-per-day limit, but users have no visibility into how many requests they've used. The only indication is in DashboardScreen.kt (line 138) which shows "Limit Reached" or a generic status — no usage count, no progress, no reset timer.

Current state in codebase

• SettingsManager.geminiApiCount: Tracks today's usage (lines 105-108)
• SettingsManager.lastApiDate: Tracks which day the count is for (lines 111-114)
• SettingsManager.canUseGeminiApi(): Returns boolean — over/under limit (lines 117-123)
• SettingsManager.tryUseGeminiApi(): Atomically checks and increments (lines 125-137)
• No getter to expose the current count publicly — geminiApiCount is a private var

Proposed Solution

Add a visible usage meter showing the user their daily AI service consumption.

Requirements

• Usage counter: Show "X / 20 requests used today" in a visible location
• Progress bar: Visual indicator (green → yellow → red as limit approaches)
• Reset timer: Show when quota resets ("Resets at midnight")
• Dashboard card: Add or update the Gemini status in DashboardScreen.kt (currently just shows "Active"/"Quota Exceeded"/"Not Configured" on lines 130-142)
• Settings card: Add usage stats in the Settings API configuration section
• Future-proof: Design to work with multiple providers from issue [Feature]: Support multiple AI service providers (beyond Gemini) #35 — each provider would have its own usage tracker

Implementation notes

• Expose current count from SettingsManager (add a getter for geminiApiCount)
• Add a simple progress composable: ${usedCount} / ${dailyLimit} with a LinearProgressIndicator
• Calculate time until next reset from current time to midnight
• For Gemini, the daily limit is hardcoded at 20 — make it configurable or expose it