ProntoPR
ProntoPR - Ship Simple Features
Minimal but extensible prototype demonstrating how a team of specialised autonomous agents can safely ship incremental features to production using GitHub pull-requests, code review, automated tests, and retrieval-augmented generation (RAG).
Quick start
# Create virtual environment
python -m venv .venv
source .venv/bin/activate
# Install
pip install -e .
# Run (dry-run; does NOT touch GitHub/Docker)
ship-simple-features "Add fibonacci function to utils module" --dry-run
Alternatively, run the standalone demo script:
```bash
python examples/demo.py "Add fibonacci function to utils module" --dry-run
Set OPENAI_API_KEY in your environment to enable LLM calls.
Architecture
See docs/architecture.md for a Mermaid diagram and detailed explanation.
Configuration
| Variable | Description |
|---|---|
OPENAI_API_KEY |
API key for OpenAI Chat completions |
GITHUB_TOKEN |
Personal access token for GitHub API (if using real adapter) |
GITHUB_REPOSITORY |
owner/repo slug for PR creation |
Development
# Install dev tools
pip install -r requirements-dev.txt
# Run linters, type-checks, and tests
ruff check .
mypy ship_simple_features
pytest --cov
# Install pre-commit hooks
pre-commit install
License
MIT – see LICENSE.