Thesilenthowler029/tribe-v2-api

Force CPU inference by default

Render individual per-timestep brain frames instead of mosaic

Fix headless VTK rendering and HF_TOKEN fallback for gated Llama model

Fix single-axis brain plot rendering

Fix plotting error and return chart_data + agent_summary alongside brain image

Install uv so uvx whisperx subprocess works in video analysis

Upgrade base image to Python 3.11 to fix exca>=0.5.20 dependency

Update to Python 3.10 and unpin requirements to fix typing bugs

Downgrade to Python 3.9 to fix typing compatibility

Fix directory permissions for pip install

Fix Debian package name

Initial backend deployment