feat(commentary): refine rafters-critic persona and improve commentary prompt for humor

feat(media): enhance media request handling with timeout configuration and improved serialization

feat(media): update media configuration for FLUX.2 and VoxCPM2 models with enhanced details

feat(media): introduce MediaRouter and stubs for image and speech generation

fix: Update GPU configuration for NVIDIA models to use A100 instead of H200

fix: Update Python version to 3.13 for compatibility with local deploy environment

Refactor modal service and logging setup

Refactor code structure for improved readability and maintainability

feat: Handle FP8 KV cache incompatibility with snapshot models in build command

feat: Add FP8 quantization support for model serving with environment overrides

feat: Enhance deployment instructions and update web server label formatting

feat: Implement well-known typed fields for verdicts in output models

fix(modal): shorten cascade slug to fit DNS limit; add endpoint health tooling

feat: refine structured output handling with guided decoding and mode adjustments

feat(logging): implement structured JSON logging for vLLM and enhance observability

feat: enhance agent personas and improve model performance tuning

feat: Enhance model routing and deployment flexibility

fix: update Google Gemma model configuration to use instruction-tuned version

feat: add Nemotron-Cascade-14B-Thinking model configuration to catalogue and README

feat: unify model catalogue and self-hosted routing

fix: update endpoint URLs to reflect new app naming conventions

fix: update Python version to 3.13 for compatibility

feat: add OpenAPI docs and opt-in bearer auth via env-var secret

feat: add Modal model-serving layer, one app per provider