Fix backend inference logging and improvements

Update README files for HF router workflow

Improve agent chat tone and HF fallback behavior

Update HF router models and inference flow

Replace unavailable models with HF Inference API-verified models (Mistral, Llama 2, Zephyr, etc.)

Add contextual chat message generation when HF models don't return messages

Add comprehensive logging for HF API token and inference failures

Remove Groq models; switch registry and clients to HF-only curated list

Add Groq->HF fallback mapping and accept HUGGINGFACE_API_TOKEN

Add fire status ETA event, probabilistic lethality by intensity, display ETA in UI

Add 15+ free HF models, expand model selection, add debug logging

Trigger HF Space pod rebuild

Keep both Groq + HF models, remove rate-limited Groq models, unify model selection UI without backend labels

Add complete YAML front matter with sdk_version and python_version to READMEs

Tune fire extinguishing: inverse scaling so 2 agents = fast game (90s), more agents = slower game (110s+)

Fix place-fire range errors and improve error detail

Use model-authored dynamic chat with recent radio context

Make simulation chat lively and celebrate fire reduction

Prepare backend for HF Space and set Node 22 for Vercel

inital build