Installation

Requirements

Python 3.10+
pip (or any Python package manager)

Install from Source

git clone https://github.com/your-org/autoflow.git
cd autoflow
pip install -e ".[dev]"

The [dev] extra includes test dependencies (pytest, etc.). The base install is sufficient for running the pipeline.

Virtual Environment (Recommended)

python -m venv .venv
source .venv/bin/activate   # macOS/Linux
pip install -e ".[dev]"

LLM Provider Setup

AutoFlow defaults to the mock provider, which returns deterministic canned responses for testing. To use a real LLM, set the provider and its API key:

export OPENAI_API_KEY="sk-..."
python -m autoflow "Your request" --provider openai

export ANTHROPIC_API_KEY="sk-ant-..."
python -m autoflow "Your request" --provider anthropic

export GOOGLE_API_KEY="..."
python -m autoflow "Your request" --provider gemini

# Ensure Ollama is running locally
python -m autoflow "Your request" --provider ollama

# No API key needed — this is the default
python -m autoflow "Your request"

The mock provider returns pre-built responses for each pipeline stage, making it ideal for development and testing.

Verify Installation

# Run the test suite (271 unit tests, no API keys needed)
pytest tests/ -v -m "not integration"

# Run a quick pipeline test
python -m autoflow "Classify support tickets by priority"

All unit tests use the mock LLM provider and require no external dependencies.