How glanderness/beeftv works
BeefTV is a local-first, AI-native video creation workspace centered on a connected canvas. The main workflow is to organize ideas and reference media in a project, send prompts or media through configured model channels and creative tools, then reuse and refine the resulting text, images, video, audio, or storyboards. The graph also captures documented project/asset management, plugin extensibility, and AI art critique. README statements support the broad product flow; sampled source gives direct evidence for critique-stage calls, while most backend wiring was not sampled.
Main entry points: infinite-canvas.tsx Creative canvas · service.go Async tasks · pipeline.ts Art critique · routes.go Backend API
Components 19
Canvas creation
- Creative canvas
web/src/components/canvas/infinite-canvas.tsx - Canvas connections
web/src/pages/canvas/use-canvas-connection-controller.ts - Art critique
web/src/lib/art-critique/pipeline.ts - Critique review
web/src/lib/art-critique/review.ts - Critique rubrics
web/src/lib/art-critique/rubrics.ts - Critique annotations
web/src/lib/art-critique/annotation.ts
Generation runtime
- Async tasks
backend/internal/task/service.go - Generation registry
backend/internal/generation/registry.go - Model routing
backend/internal/app/model_router.go - Model providers
backend/internal/provider/registry.go
Workspace data
- Projects and workflow
backend/internal/project/service.go - Asset library
backend/internal/asset/service.go - Workspace storage
backend/internal/repository/repository.go
Tools and extensions
- Provider protocols
backend/internal/protocol/registry.go - Plugins
web/src/lib/plugins/plugin-types.ts
App runtime
- Backend API
backend/internal/handler/routes.go - Desktop runtime
backend/cmd/desktop/app.go
Other components
- Creator
- Configured AI service
Connections 19
- Creator → Creative canvas: creates in
README.md - Creative canvas → Canvas connections: organizes with
README.md - Creative canvas → Projects and workflow: works in
README.md - Creative canvas → Art critique: requests critique
README.md - Art critique → Critique rubrics: builds prompts
web/src/lib/art-critique/pipeline.ts - Critique review → Critique annotations: validates targets
web/src/lib/art-critique/review.ts - Art critique → Configured AI service: requests analysis
web/src/lib/art-critique/pipeline.ts - Creative canvas → Async tasks: starts generation
README.md - Async tasks → Model routing: uses channels
README.md - Model routing → Model providers: routes requests
- Model providers → Provider protocols: adapts protocols
- Async tasks → Asset library: produces media
README.md - Projects and workflow → Workspace storage: persists workspace
README.md - Asset library → Workspace storage: stores assets
README.md - Creative canvas → Plugins: uses extensions
- Creator → Desktop runtime: opens workspace
README.md - Desktop runtime → Backend API: serves workspace
- Backend API → Projects and workflow: handles project actions
- Backend API → Async tasks: handles task actions