[Feature]: Show token usage and rate-limit warnings per turn

[shauryagangrade]

Problem

Free OpenRouter models are rate-limited and users get no feedback on how
much context they're consuming. ChatOpenAI streams carry usage metadata,
but GCode drops it.

Proposed approach

• Capture usage/rate-limit info from stream chunks (response_metadata,
  OpenRouter X-RateLimit-* headers).
• Print a compact footer after each turn: tokens in/out, approximate cost
  (OpenRouter publishes per-token pricing), and remaining rate limit.
• Add a /usage command or fold it into /history.

Where to look

• gcode/agent.py:59-79 (_stream), :101-137 (run_turn)
• gcode/ui.py for footer rendering hooks.

Acceptance criteria

• A turn prints usage when the provider supplies it; silently omits otherwise.
• Rate-limit warnings appear before a 429 aborts the request.

Difficulty

Medium.