Manage environments and packages#

Create a regular environment#

Run Python Envs: Create Environment, choose Conda Code, then choose one of the regular environment types:

Named environment: Creates an environment by name in one of the environment directories configured for the primary conda.
Project prefix: Creates an environment at .conda below the registered project. This secondary option is interactive only.

Conda Code includes Python unless the requested package list already contains a Python specification.

Create from a project file#

Use Create from project files for environment.yml, explicit.txt, and supported conda-lock inputs. These create regular named environments.

Select an environment#

Use Set As Project Environment in the Python Environments view. A workspace environment can only be selected for the project that declares it. Regular environments can be selected for any compatible project.

Conda Code persists the manager-wide and project selections in the current VS Code workspace.

Remove an environment#

Use Delete Environment in the Python Environments view.

Conda Code removes:

named conda environments
project prefixes located exactly at <project>/.conda

Conda Code routes removal through the prefix’s owning conda installation. It refuses to remove the base environment, another conda installation, a prefix that it cannot identify as named or <project>/.conda, a prefix without a usable owner executable, or a prefix without valid conda-meta data.

Manage packages#

Open Manage Packages below a Conda Code environment.

For regular environments, Conda Code can:

install packages
update requested packages
remove packages
list conda package records

Package changes use the prefix’s owner executable. Conda Code keeps a prefix without a usable owner executable visible but refuses to change it.

Raw pip-only distributions are omitted. Packages installed by conda-pypi have conda records and remain visible.

See the Package operations for the workspace differences.

Export an environment SBOM#

Install conda-sboms 0.3.0 or newer from conda-forge by running this through the routed conda executable:
```
conda install --name base --channel conda-forge "conda-sboms>=0.3.0"
```
Replace conda with cx, cxz, or the configured executable path when needed. The installation guide covers alternative methods.
Run the routed executable with export --help and confirm that cyclonedx-json-v1.7 is listed.
Select the Conda Code environment for the active editor scope.
Run Conda Code: Export Environment SBOM and choose the JSON destination.

The save dialog proposes <environment>.cdx.json. Conda Code exports with --from-history, which preserves requested package roots when conda history contains them while the SBOM still includes the resolved package set.

For a regular environment, install the plugin beside the conda executable that owns its prefix. For a workspace environment, install it beside the configured primary conda executable.