docs: explain AI-native libraries and automatic execution logs

[philipgreat]

Summary

• Explain why TeaQL is an AI-native library and runtime: business-shaped Q/E/mutation APIs, progressive model-aware Assist, software checks and a smaller application review surface.
• Make automatic execution logs without handwritten logger calls a core debugging and operations benefit, with language-native intent declarations and existing masking/sink boundaries.
• Link the Robot Task Board showcase and keep native examples, installation, published versions and supported scope.
• State explicitly that token savings have not been measured in a controlled comparison.

Validation

• README-only diff; no runtime, generated library, dependency or version changes.
• Whitespace, local links, heading anchors and per-language Assist prefixes checked.
• Native syntax checks completed for Rust, Go, TypeScript, Swift and Python excerpts during the rewrite; Java/.NET retain existing API calls with named child requests.
• The automatic-log follow-up changes prose only; code/log examples and the installation/scope/safety tail remain unchanged.
• Runtime/example CI is checked separately before merging. No new runtime release is requested.