Create from project files#

Conda Code recognizes these files at a registered Python project root:

  • environment.yml

  • environment.yaml

  • explicit.txt

  • conda-lock.yml

  • conda-lock.yaml

An existing discovered conda workspace takes precedence. Creating an environment for that project installs one of its declared environments instead.

Use environment.yml#

  1. Put environment.yml or environment.yaml at the project root.

  2. Run Python Envs: Create Environment.

  3. Select the file.

  4. Enter the environment name.

Quick Create skips the file and name prompts when there is one recognized input. It derives an available name from the project directory.

Conda Code runs this command from the project root:

conda create --name NAME --file environment.yml

The command-line name overrides name or prefix in the CEP 24 file. Packages selected during creation are installed after the file-based environment is created.

The file is not watched or updated. Use regular package management for the created prefix.

Use explicit.txt#

Place a valid CEP 23 explicit file at the project root. It must contain the @EXPLICIT marker.

Create the environment through Quick Create or select explicit.txt during interactive creation. Conda Code passes --no-default-packages so configured creation defaults cannot change the exact package set.

Do not select additional packages during creation. Conda Code refuses the operation instead of changing the exact input.

Use a conda-lock file#

Install conda-lockfiles 0.2 or newer in the configured conda base environment. Follow the conda-lockfiles installation instructions using the channels configured for your conda distribution.

Put conda-lock.yml or conda-lock.yaml at the project root, then create the environment through Quick Create or interactive creation.

Conda Code treats these as exact inputs. It disables configured default packages and refuses additional creation packages.

Choose among several files#

Quick Create fails when the project root contains several recognized inputs. Run Python Envs: Create Environment and choose the file to use.

The same interactive menu also offers a conda workspace, a project .conda prefix, and a regular named environment.