Skip to content
vebbaybiPublic

About

AI-powered video intelligence system built specifically for streamers, content creators, and editors who work with large sets of reaction videos, clip dumps, gameplay captures, or long-form session recordings.

Resources

Contributing

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

ClipSense

ClipSense is an early AI-assisted video pre-editing prototype. The intended product will eventually include a desktop processing workstation and a web workspace, but this repository currently contains only a partial web MVP: a Next.js browser app, a Go HTTP API, a Python processing worker, and local infrastructure definitions.

The repository is in a controlled re-foundation. Start with Current State before attempting runtime work.

Physical Runtime

Component Path Current purpose
Web apps/web Login, ZIP submission, batch list/detail UI
API apps/api Auth, batches, upload, retrieval, export
Processor apps/api/ai_worker ZIP extraction, transcription, analysis, vectors
Local stack docker-compose.yml Postgres, Redis, Qdrant, API, worker, web

There is no desktop application, shared package workspace, OpenAPI contract, versioned migration system, or end-to-end test suite yet. Extensionless files that show directory trees are historical blueprints, not packages or applications.

Local Runtime

The root docker-compose.yml is the authoritative local stack definition. On a machine with Docker and Docker Compose:

docker compose config
docker compose build --no-cache
docker compose up -d
docker compose ps

The web application is served at http://localhost:3000. The public landing page is /; authenticated application routes begin at /dashboard. The API exposes:

  • GET http://localhost:8080/api/health/live for process liveness.
  • GET http://localhost:8080/api/health/ready for Postgres and Redis readiness.
  • GET http://localhost:8080/api/health as a compatibility alias for readiness.

Stop the stack normally with:

docker compose down

The authoritative Sprint 1 Linux container gate is the docker-runtime job in .github/workflows/ci.yml. It builds the application images, starts the canonical Compose stack, checks dependency and route behavior, verifies readiness degradation/recovery and graceful termination, and retains commit-attributed diagnostics. Docker Desktop on the HP or Acer is optional supplemental Windows QA. Future desktop application testing remains a separate cross-platform concern.

The current Compose defaults are for local development only. In particular, production-sensitive secret and port hardening remains Sprint 1 Work Unit 2 work.

Current Limitations

The Docker MVP is not yet demonstrated end to end until the authoritative GitHub Actions runtime gate passes. Full media processing and cold model acquisition are outside that boot/integration gate and may require a future manual or scheduled smoke. Authenticated CSV export is broken, the upload cap is ineffective, and Compose does not provide a secure JWT secret; those remain Work Unit 2 concerns.

Inspect And Validate

# Repository/environment report
powershell -NoProfile -ExecutionPolicy Bypass -File scripts/doctor.ps1

# Go unit tests
Set-Location apps/api
go test ./...

# Python syntax and ZIP-safety tests
Set-Location ../api/ai_worker
python -m py_compile main.py zip_safety.py
python -m unittest discover -s tests

# Web build
Set-Location ../../../web
npm ci
npm run build

Docker commands require Docker Engine and Compose:

docker compose config
docker compose up --build

Do not interpret a successful Compose configuration parse as proof that containers start or that the product workflow succeeds.

Repository Truth

About

AI-powered video intelligence system built specifically for streamers, content creators, and editors who work with large sets of reaction videos, clip dumps, gameplay captures, or long-form session recordings.

Resources

Contributing

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages