You're probably very familiar with the tools you use daily and operate them from muscle memory. But there are also these setup or maintenance tasks you only do every X months and their practical details are a bit hazy.

This is a random, work-in-progress collection of cheat sheets for these commands or workflows I use too infrequently to remember.


  • List installed Jupyter/JupyterLab kernels (run from env that has jupyter installed):

     jupyter kernelspec list
  • Add a new kernel for a custom virtual/conda env:

    • Enter the virtual/env (e.g. . bin/venv/activate or conda activate $envname)
    • Install ipykernel package: e.g. pip install ipykernel
    • Create kernel spec for jupyter/jupyterlab:
      python -m ipykernel install --user --name $kernelname

    Optionally add --display-name "Pretty Name" to give it a prettier name in the Jupyter(Lab) UI - Check that kernel is now listed at jupyter kernelspec list (run this from env where you run Jupyter(Lab))


  • Updating
    • update conda (base environment): conda update conda
    • Update all packages in an environment: conda update --all
  • Remove unused packages and caches (can save up multiple GB).

     conda clean --all

    (optionally add --dry-run)

  • Remove environment conda env remove -n $ENVNAME


  • List installed formulas
  • Including dependencies: brew list
  • Only the "leave" formulas: brew leaves
  • How do I update my local packages?

     # First update the formulae and Homebrew itself:
     brew update
     # You can now find out what is outdated with:
     brew outdated
     # Upgrade everything with:
     brew upgrade
     # Or upgrade a specific formula with:
     brew upgrade <formula>
  • By default, Homebrew does not uninstall old versions of a formula. To clean up the accumulation of old, unused versions:

     brew cleanup <formula>
     # or clean up everything at once (add `-n` for dry run):
     brew cleanup
  • Remove packages that were once installed as dependency of another, but not necessary anymore:

     brew autoremove
  • General checkup

     brew doctor


  • First, it's best to clean up all stopped containers (assuming these are all just disposable leftovers).

    # Check first for stopped containers
    docker ps -a
    # Prune stopped containers
    docker container prune
  • Clean up dangling images

    # List images we currently have
    docker image ls
    # Prune dangling images (untagged and not being used by a container)
    docker image prune
  • A bit more aggressive: prune all images (also tagged ones) that were created long enough ago

    docker image prune -a --filter "until=2022-01-01"


(yes, I use vim too infrequently to remember these, sue me)

  • Change a word (e.g. interactive git rebase): :cw
  • show line numbers :set number