VSCode + WSL + Conda Setup (with unittest + Tasks)

Posted on February 13, 2026 · 6 mins read

VSCode + WSL + Conda Setup (with unittest + Tasks)

This post documents a reproducible setup for developing a Python project inside WSL using VS Code, with a dedicated Conda environment, unittest discovery, and convenient VSCode Tasks / Debug configs. This setup is for developing drafts-checker.py - Stage-Based Preparation Tool for Publishing Recipes

  • https://github.com/sagivba/family_recipes/blob/main/code/drafts-checker.py
  • https://sagivba.github.io/family_recipes/

What .vscode/ is

.vscode/ is a project-local configuration folder for Visual Studio Code. Files here apply only to the current repository/workspace (not global machine settings). It’s the place to store reproducible editor/run/debug/test settings for the project.

Files we create inside .vscode/

  • .vscode/settings.json: Workspace settings (interpreter, test discovery, env files, terminal behavior)
  • .vscode/launch.json: Debug configurations (what runs when you press F5)
  • .vscode/tasks.json: Reusable command shortcuts (Tasks: Run Task) such as tests, checkers, docker compose

0) Prerequisites

  • Windows with WSL2 (Ubuntu)
  • VS Code installed on Windows
  • Miniconda/Conda installed inside WSL (paths like /home/<user>/miniconda3/...)

1) Install the WSL Remote Extension in VS Code

In VS Code:

  • Extensions -> install WSL (Microsoft) / Remote - WSL

Then open the project as a WSL workspace (not Windows paths).

From a WSL terminal:

cd ~/src/family_recipes
code .

This ensures VS Code runs the workspace in the WSL context (so it can see Linux interpreters and Linux conda envs).

3) Create and Activate a Dedicated Conda Environment

Inside WSL:

conda create -n family_recipes python=3.12 -y
conda activate family_recipes
which python
python -c "import sys; print(sys.executable)"

Expected:

  • which python and sys.executable point to: /home/<user>/miniconda3/envs/family_recipes/bin/python

4) Install Dependencies (Conda + pip)

Some dependencies are best installed via conda, while others are typically installed via pip.

Install via conda (examples)

conda activate family_recipes
conda install -y click pyyaml
python -c "import click, yaml; print('ok')"

Install via pip (OpenAI + dotenv)

conda activate family_recipes
python -m pip install -U pip
pip install openai python-dotenv
python -c "import openai; from dotenv import load_dotenv; print('ok')"

5) Run Unit Tests from CLI (unittest discover)

From the repo root:

python -m unittest discover -s code/_Tests -p "test_*.py" -v

6) Set Project Environment Variables (PYTHONPATH)

Create a repo-level .env file (used by VS Code Python tooling):

<repo>/.env

PYTHONPATH=code

7) Configure VS Code (Workspace Settings)

Create:

<repo>/.vscode/settings.json

{
  "python.defaultInterpreterPath": "/home/sagivba-adm/miniconda3/envs/family_recipes/bin/python",
  "python.testing.unittestEnabled": true,
  "python.testing.pytestEnabled": false,
  "python.testing.unittestArgs": [ "-v", "-s", "code/_Tests", "-p", "test_*.py" ],
  "python.envFile": "${workspaceFolder}/.env",
  "python.terminal.activateEnvironment": true
}

Notes:

  • The file must contain one JSON object (not multiple {...}{...} blocks).
  • After editing, run: Developer: Reload Window.

8) Configure Debug (unittest)

Create:

<repo>/.vscode/launch.json (example)

{
  "version": "0.2.0",
  "configurations": [
    {
      "name": "unittest: all (discover)",
      "type": "python",
      "request": "launch",
      "module": "unittest",
      "args": ["discover", "-s", "code/_Tests", "-p", "test_*.py", "-v"],
      "console": "integratedTerminal",
      "justMyCode": true,
      "envFile": "${workspaceFolder}/.env"
    }
  ]
}

9) Configure Tasks (Run Common Commands in One Click)

Maintain:

<repo>/.vscode/tasks.json

Key points:

  • Use ${command:python.interpreterPath} so Tasks always run with the interpreter selected in VS Code (your conda env).
  • Create tasks for:
    • Docker compose (dev start/stop/logs)
    • unittest discover
    • drafts-checker.py (with and without --use-ai)
    • A “CI quick” meta-task that runs tests + checker in sequence

10) OpenAI API Key (for --use-ai mode)

To run the pipeline with AI features, export your key:

export OPENAI_API_KEY="YOUR_KEY_HERE"

Persist it (one common approach) by adding to ~/.bashrc:

echo 'export OPENAI_API_KEY="YOUR_KEY_HERE"' >> ~/.bashrc
source ~/.bashrc

Verify:

echo "$OPENAI_API_KEY" | head -c 8; echo

11) Sanity Check (VS Code + WSL + Env)

Inside VS Code terminal:

conda activate family_recipes
which python
python -c "import sys; print(sys.executable)"
python -m unittest discover -s code/_Tests -p "test_*.py" -v

Expected:

  • Python path points to the conda env
  • Tests run and pass
← Previous Post