Installation
Requirements
- Python 3.11 or higher
- A project with docstrings, Pydantic models, or PySpark schemas
No mutation testing tool required. Quell reads specs, not mutation results.
Install
Create a virtual environment
python -m venv .venv
# macOS / Linux
source .venv/bin/activate
# Windows
.venv\Scripts\activate
Or with uv (faster):
uv venv && uv pip install quelltest
Install quelltest
pip install quelltest
The PyPI package is quelltest. The CLI command is quell. They're different names — this is intentional.
Verify the install
quell --version
Expected output:
quelltest 0.5.0
Authenticate (one-time)
quell auth login
This opens your browser to quell.buildsbyshashank.tech/auth/login. Sign in with your Quell account, click Authorize CLI, and return to the terminal. Your credentials are saved to ~/.quell/credentials.json.
No account yet? Sign up free — no credit card required. The Hobby plan covers 500 requirements/month with the full rule engine, no API key needed.
To check who you're logged in as:
quell auth status
To log out:
quell auth logout
Optional extras
PySpark schema scanning
pip install "quelltest[pyspark]"
Required only if your project uses PySpark StructType schemas. The base install already handles docstrings and Pydantic models.
LLM fallback
The rule engine handles ~75% of requirements with no network call. For the rest, configure an LLM provider:
export ANTHROPIC_API_KEY=sk-ant-... # Claude
export OPENAI_API_KEY=sk-... # GPT-4
Or add to pyproject.toml:
[tool.quell]
llm_provider = "anthropic"
--no-llm is the default. You have to explicitly configure a provider to enable LLM fallback. Without one, complex requirements are flagged but not sent anywhere.
CI environments
In CI, skip quell auth login and use an API key instead:
export QUELL_API_KEY=qk_live_...
quell check src/ --no-llm --ci --threshold 0.80
Get your API key from the dashboard.
Updating
pip install --upgrade quelltest
Platform notes
Quell runs natively on macOS, Linux, and Windows. No WSL required on Windows. The CLI, verification engine, and AST spec readers all work cross-platform.