disable flaky LLM judge layer (default-denied every clean wish on llama.cpp); wordlist layers 1-2 remain the gate

security: re-moderate model text, identity-bound rite token, XFF/queue-freeze hardening, OOM guard; README truth fixes

inline ticker flow, forgiving water args, wish-counter from traces, canonical namespaced tags

planet owns the root (evict gradio SPA index)

gradio Server mode + browser-invoked granting (the her idiom)

preload libcudart for mamba kernel import

GODSEED v0