Add framework-level OpenTelemetry tracing

[cpsievert]

Summary

Chatlas now emits OpenTelemetry spans that capture the full structure of multi-turn conversations and tool execution — without requiring any provider-specific instrumentor libraries. When a TracerProvider is configured, every chat()/stream() call automatically produces a 3-level span hierarchy:

invoke_agent                      # wraps the full chat loop
├── chat gpt-4o                   # each model API call
├── execute_tool get_weather      # each tool invocation
├── chat gpt-4o                   # follow-up model call
└── ...

Users opt in with pip install "chatlas[otel]" and a standard TracerProvider setup (console exporter, Logfire, or any OTLP-compatible backend). The approach is consistent with Shiny for Python's OTel story — same [otel] extra pattern, same recommended tools, same config-module pattern.

Spans follow the GenAI semantic conventions and record token usage, response model/ID, and optionally full message content (gated by OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT). These framework spans complement (not replace) provider-specific SDK instrumentors like opentelemetry-instrumentation-openai-v2.

• New chatlas/_otel.py module with span lifecycle functions
• Hooks into all 6 core Chat methods (sync + async for agent/chat/tool spans)
• 7 tests with VCR cassettes covering span hierarchy, token usage, content capture, tool errors, streaming, and no-op behavior
• Updated docs/get-started/monitor.qmd with framework-level tracing docs

Test plan

• pytest tests/test_otel.py — 7/7 passing
• pyright chatlas/_otel.py — 0 errors
• ruff check and ruff format — clean
• Manual verification with a real exporter (Logfire or console) and live API key