Ship and recap

Final step. The workshop ships with a Dockerfile so the FastAPI inspector runs anywhere. The MCP server still runs locally over stdio for desktop clients. Same image, two entrypoints, zero drift.

Dockerfile

dockerfile

FROM python:3.11-slim
COPY --from=ghcr.io/astral-sh/uv:latest /uv /uvx /bin/

WORKDIR /app

RUN apt-get update && apt-get install -y curl \
    && rm -rf /var/lib/apt/lists/*

COPY pyproject.toml ./
RUN uv sync --frozen

COPY . .

ENV PYTHONUNBUFFERED=1
ENV PATH="/app/.venv/bin:$PATH"

EXPOSE 8000

CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "8000"]

Small image, uv for fast dependency resolution, and a direct uvicorn launch for the inspector. MCP stdio runs against the same container (make mcp) when you need it.

Ordering exercise: Order the full pipeline end to end

Loading practice…

Look what you shipped. A production-shaped MCP server with three working tools, structured review output, GitHub write-back, content-hash idempotency, and a FastAPI inspector on the same service layer. Any MCP client can now drive an end-to-end PR review against your GitHub account.

Quiz: Quiz

Loading practice…