Conda Code

Conda environments, workspaces, packages, and tasks for Visual Studio Code.

Conda Code adds its own conda environment and package manager to the Python Environments extension. It discovers regular environments across local conda installations and integrates conda-workspaces workspace environments and tasks in the same provider.

Run Conda Code: Create Environment from File from environment.yml. The resulting regular named environment appears as the selected project environment in the shared Python Environments UI. Python Environments supplies the VS Code views. Conda Code supplies the Conda Code manager and performs conda discovery, creation, activation, package changes, and removal.

Get started

1. Install Conda Code from the Visual Studio Marketplace or run:

   code --install-extension jezdez.conda-code
   

   Python Environments is installed as an extension dependency.

2. Make Conda Code the default conda manager in your user or workspace settings:

   {
     "python-envs.defaultEnvManager": "jezdez.conda-code:conda",
     "python-envs.defaultPackageManager": "jezdez.conda-code:conda"
   }
   

3. Open the Python Environments view and choose Conda Code.

Note: Python Environments currently registers its built-in conda provider unconditionally and does not deduplicate environments across providers. Both conda branches can appear. The settings above select Conda Code for creation and package operations.

Follow the quickstart to check regular environment and workspace discovery.

What Conda Code handles

• Regular environments: Find base, named, and prefix environments without starting their Python interpreters, including environments without Python.
• Creation inputs: Create named environments from package specifications, environment.yml, CEP 23 explicit files, and conda-lockfiles.
• conda-workspaces: Discover, install, activate, and manage workspace environments, then run declared tasks from the manifest editor through native VS Code Tasks. Shell activation targets the exact installed workspace prefix. Conda Code detects richer workspace snapshots and declaration locations when the installed conda-workspaces provides them.
• Owner-aware operations: Route activation, package changes, and safe deletion through the conda installation that owns each prefix. Run Conda Code: Export Environment SBOM to use that owner for a regular environment or the configured primary conda installation for a workspace.
• Ecosystem compatibility: Include conda-pypi package records, ignore conda-global tool prefixes and conda-exec cache prefixes, and leave Pixi workspaces to Pixi Code when it is installed.

Run Conda Code: Refresh Environments after an external change that is not picked up automatically.

Requirements

Required:

• Visual Studio Code 1.118 or newer
• conda 26.3 or newer
• A trusted VS Code window for project operations
• A local folder registered as a Python project for project-scoped creation and workspace discovery

Optional integrations:

• conda-workspaces 0.7 or newer for workspace environments and tasks
• conda-lockfiles 0.2 or newer for conda lockfile creation inputs
• conda 26.5 or newer for workspace PyPI dependencies
• conda-sboms 0.3.0 or newer for selected-environment CycloneDX export

Newer conda-workspaces capabilities are detected at runtime. The 0.7 integration remains available when complete snapshots or dependency declaration locations are not reported.

Set conda-code.condaExecutable when the primary conda is not available through CONDA_EXE, the Python extension's python.condaPath setting, or PATH.

See the complete requirements for installation details.

Learn more

• Create environments from project files
• Use conda-workspaces
• Understand the provider model
• Troubleshoot Conda Code
• Read the changelog

Report problems and request features in the issue tracker.

Development

Use the Node.js release pinned in .nvmrc.

npm ci
npm run typecheck
npm test
npm run docs

Run the Extension launch configuration in an Extension Development Host. Build an installable package with npm run vsix.

License

Conda Code is distributed under the BSD 3-Clause License.