Hello everyone, I have a flow and multiple tasks that have to run in a specific conda environment. For this I need to use a Docker Worker, but because of the infrastructure that this has to be deployed in, I am not able to create custom docker images. Is there a way to use a Prefect Conda Docker Image and update the environment with a specific environment file during the pull step? I have tried to use

prefect.deployments.steps.run_shell_script

where I run

conda env update --file <path_to_environment_file>

to install the required packages. This all works, but when the Flow is now executed in this docker container the installed packages aren't found.

hi @Steffen - what base image are you using? for example there are tags like

prefecthq/prefect:3.2.7-python3.12-conda

on dockerhub with `uv` pip or conda (most easily with uv) you can do stuff like this to install things at runtime

Hi @Nate - Thank you for your response. Yes, I've been using

prefecthq/prefect:3.2.7-python3.12-conda

as my base image. I'm currently working on some other stuff but I'll try out your recommendation in the upcoming days.