Detect ZeroGPU from Spaces environment

Force local ZeroGPU inference path on Space

Fix Gradio entrypoint and switch explainability to Gemini

Serve frontend API routes via Gradio Server

Mount Gradio app at Space root

Fix ZeroGPU model loading lifecycle

Fix ZeroGPU startup and local GPU inference path

Mount FastAPI routes into Gradio space app

align project with CLAUDE spec and hf space deploy

fix(video): cap frames at 60, pass tmp_path to coherence, keep temp alive through inference

feat: Enhance generator detection and attribution mechanisms

feat: add metadata extraction and synthetic keyword detection for images and videos

fix: update Dockerfile dependencies, remove audio processing, and replace models

feat: enhance model capabilities and update dependencies

Update model stack, add audio stream extraction, and harden inference wiring

Initial commit