Thanks for the details — the error tells us exactly what’s happening: Docker is trying to start your flow container by running

prefect ...

, but there’s no

prefect

executable on the container’s PATH at runtime: exec: "prefect": executable file not found in $PATH Why this happens with your Dockerfile -

uv sync

installs your project (including Prefect, if listed) into a virtual environment under

/app/.venv

. - You verify Prefect during build with

uv run prefect --version

(which runs inside that venv), but at runtime the Docker worker runs

prefect

directly — not via

uv run

. - Because

/app/.venv/bin

is not on PATH, the

prefect

CLI isn’t found when the container starts. Two solid ways to fix it A) Keep using uv + project venv (recommended) - Make sure Prefect is in your pyproject dependencies. - Put the venv’s bin directory on PATH so

prefect

is discoverable at runtime. Example Dockerfile:

FROM python:3.12-slim

# System deps
RUN apt-get update && apt-get install -y --no-install-recommends \
    curl \
    ca-certificates \
    libgomp1 \
    build-essential \
    gcc \
    libpq-dev \
  && rm -rf /var/lib/apt/lists/*

# Install uv
ADD <https://astral.sh/uv/install.sh> /uv-installer.sh
RUN sh /uv-installer.sh && rm /uv-installer.sh
ENV PATH="/root/.local/bin:$PATH"

WORKDIR /app

# Install project deps (including prefect) into /app/.venv
COPY pyproject.toml uv.lock ./
RUN uv sync --frozen

# Make the project venv active for all runtime processes
ENV VIRTUAL_ENV=/app/.venv
ENV PATH="${VIRTUAL_ENV}/bin:${PATH}"

# Optional: set PYTHONPATH to your app dir
ENV PYTHONPATH="/app:${PYTHONPATH}"

# Copy the rest of your code
COPY . /app

# (Optional) Quick smoke test
RUN which prefect && prefect --version

Key points: - Ensure

prefect

is listed in your

pyproject.toml

dependencies. - The

ENV PATH="${VIRTUAL_ENV}/bin:${PATH}"

line is what makes

prefect

available at runtime. B) Install Prefect into the system Python instead - If you prefer not to rely on a venv at runtime, install Prefect system-wide so

prefect

is on PATH without extra env changes:

# after setting up Python
RUN pip install --no-cache-dir 'prefect>=3.0.0'
# or with uv:
# RUN uv pip install --system 'prefect>=3.0.0'

If you go this route, also install your project dependencies system-wide instead of to a venv, or the Prefect process won’t see your project packages. Small corrections to your Dockerfile - This line is incorrect and creates an env var literally named "export PYTHONPATH":

ENV export PYTHONPATH=$PYTHONPATH:.

Use:

ENV PYTHONPATH="/app:${PYTHONPATH}"

- Consider using a supported Python version (e.g., 3.11 or 3.12). 3.14 may not be fully supported by all dependencies yet. Optional quick workaround (without rebuilding) - If you know the venv lives at

/app/.venv

, you can add its

bin

directory to PATH via your Docker work pool’s environment: - In the Docker work pool settings (Base job template) or via deployment

job_variables

, set:

"env": {
      "PATH": "/app/.venv/bin:${PATH}"
    }

This works only if your image already has

/app/.venv/bin/prefect

present. Other notes - You’re using

push=False

. Make sure the Docker worker is running on the same host that built the image so it can find

wwwaste-data-pipeline:v2

. If the worker is remote, set

push=True

and use a registry. Would you share: - Your

pyproject.toml

dependencies (to confirm

prefect

is included)? - Where your Docker worker is running (same machine vs remote)? With that, I can tailor the exact Dockerfile and deployment settings.