Troubleshooting VS Code
Python interpreter not found
The project venv is created by uv at .venv/. The interpreter path is set in
.vscode/settings.json:
/workspaces/teamster/.venv/bin/python
If VS Code is not picking it up, open the command palette and run Python:
Select Interpreter, then choose the .venv path above.
If the venv doesn't exist yet:
uv sync --frozen
Import errors / Pylance red underlines
Pylance resolves imports from the venv. If imports are unresolved after
uv sync:
- Reload the VS Code window (Developer: Reload Window).
- Confirm the interpreter is set to
.venv(see above).
Type checking is intentionally set to off in .vscode/settings.json — Pyright
via Trunk handles type checking at lint/commit time instead.
Secrets not loaded
Environment variables are fetched on demand from 1Password by
tests/conftest.py when pytest runs. For commands outside pytest (e.g.,
dagster definitions validate), run them in the VS Code terminal where
OP_SERVICE_ACCOUNT_TOKEN is available.
Trunk linter not running
Trunk is the default formatter for Python, SQL, YAML, and Markdown
("editor.defaultFormatter": "trunk.io"). If format-on-save is not working:
- Confirm the Trunk extension is installed and enabled.
- Check that
./trunkis executable:chmod +x /workspaces/teamster/trunk ./trunk install - Check the Trunk output panel (View → Output → Trunk) for errors.
Container needs a rebuild
If postCreate did not run completely, the environment may be partially set up.
Re-run the setup steps manually:
uv sync --frozen
Or rebuild the container from scratch: open the command palette and run Dev
Containers: Rebuild Container. Note that a full rebuild re-runs dbt deps and
dbt parse for all 15 projects, which takes several minutes.
dbt Power User
See Troubleshooting dbt → dbt Power User.
See also: Guides · Troubleshooting dbt · Troubleshooting Dagster