stash-ai-server 0.1.1a18

Stash AI Overhaul Backend Server

AI Overhaul Backend

Minimal FastAPI backend with SQLite + Alembic migrations.

Features

• FastAPI application exposing /api/v1/requests CRUD (create + read/list now)
• SQLite database stored under backend/data/app.db
• SQLAlchemy 2.0 style models
• Alembic migration management
• Docker dev environment with live reload (code mounted as volume)

Directory Layout

backend/
  stash_ai_server/
    api/          # Routers
    core/         # Config
    db/           # Session + base
    models/       # SQLAlchemy models
    schemas/      # Pydantic schemas
    main.py       # FastAPI entry
  alembic/        # Migration scripts
  alembic.ini     # Alembic config
  requirements.txt
  Dockerfile

Run (Docker Compose)

docker compose up --build

Then visit http://localhost:4153/docs for the Swagger UI.

During development, source changes under backend/stash_ai_server auto-reload thanks to --reload and bind mount.

Manual Local Run (Without Docker)

python -m venv .venv
. .venv/Scripts/activate  # Windows PowerShell: .venv\Scripts\Activate.ps1
pip install -r backend/requirements.txt
uvicorn stash_ai_server.main:app --reload

Configuration file

For secrets and a few runtime overrides (data dir, plugin dir, stash API key), copy config.sample.env to backend/config.env and edit values. This file is gitignored so your secrets won't be checked in.

Example:

cp backend/config.sample.env backend/config.env
# edit backend/config.env and then start the server

When running in Docker Compose the image expects /app/data and the compose file or secrets file should mount the appropriate host paths.

Conda / pip install users

If you install the package from PyPI (or from the built wheel), the alembic folder and alembic.ini are included in the package so you can run migrations after creating backend/config.env. For local editable installs (pip install -e .) ensure you run from the repo root so alembic is present in the working tree.

Apply Migrations

The container runs with Base.metadata.create_all() for convenience right now. To use Alembic migration fully:

# Inside container or local env (with PYTHONPATH set to backend):
alembic upgrade head

Example Requests

POST /api/v1/requests
{
  "prompt": "Describe scenic imagery"
}

GET /api/v1/requests
GET /api/v1/requests/1

Next Steps

• Add status update endpoint (PATCH)
• Introduce background processing queue
• Add filtering/pagination on list endpoint
• Add CORS config if frontend fetches directly
• Authentication (API key or session) when needed