[tzhouam]

LingBot World’s model knowledge exposed only two lifecycle rules and lacked guidance for its implemented execution paths. Add architecture and execution/validation pages, six source-backed rules covering KV reuse, camera alignment, pure Ulysses, VAE sharding and memory, and online FP8, and refresh the model navigation.

The new guidance is pinned to vLLM-Omni 9cf443a7bb24d6b8007981311466d769ec13b3d8. Shared runtime rules are linked from the model pages, and the repository-wide release baseline is preserved.

Validation:

• Knowledge tree: 0 errors (60 existing warnings); wiki lint: 0 errors or warnings.
• vLLM-Omni release drift audit in enforce mode: clean.
• Native knowledge reader and related-page retrieval resolve the new pages and source pin.
• Independent review against pinned upstream source completed; no GPU execution results are claimed.

Refs #333. This completes the LingBot World backfill; the issue remains open for intake accounting and legacy coverage reconciliation.

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, add credits to your account and enable them for code reviews in your settings.