Docker for Python services

A container image is the unit you deploy: an interpreter, your dependencies, your code, and a start command, built once and run identically on a laptop, in CI, and on the platform. For a uv project the Dockerfile is short, and the whole craft is in the order of the lines. Dependencies change rarely and take a minute to install; source changes many times a day and copies in a second. Put the slow, stable layer first and the fast, volatile layer last, and a rebuild after a code change takes seconds.

FROM python:3.14-slim
COPY --from=ghcr.io/astral-sh/uv:0.9 /uv /uvx /bin/
ENV UV_COMPILE_BYTECODE=1 UV_LINK_MODE=copy UV_PYTHON_DOWNLOADS=never
WORKDIR /app