Development Setup¶

Prerequisites¶

Python 3.11+
git

Install¶

git clone https://github.com/vamshisiddarth/argus.git
cd argus
python3.11 -m venv .venv && source .venv/bin/activate
pip install -e ".[dev]"
pre-commit install

Running tests¶

All tests run offline — no cloud credentials needed:

pytest tests/ -v                   # 528 unit tests
make test-integration              # 32 integration tests (adapter contracts, report schema)
make test-all                      # everything (560 tests)

Subsets:

pytest tests/adapters/aws/ -v     # AWS adapter
pytest tests/adapters/gcp/ -v     # GCP adapter
pytest tests/adapters/azure/ -v   # Azure adapter
pytest tests/ai/ -v               # AI providers
pytest tests/core/ -v             # Agent loop, chat session, report generation

With coverage:

pytest tests/ --cov=. --cov-report=term-missing

Code style¶

Pre-commit hooks run ruff (lint + format) and mypy automatically on each commit. To set them up:

pre-commit install          # one-time setup
pre-commit run --all-files  # manual run on all files

You can also run the tools directly:

ruff format .     # format
ruff check .      # lint
ruff check --fix  # auto-fix lint issues
mypy .            # type check

Rules: - Line length: 88 characters - Type hints on all public functions - No bare except Exception - Python 3.11+ minimum (match, | union types freely; avoid 3.12+ type statement)

Running the docs site locally¶

pip install mkdocs-material mkdocs-minify-plugin
mkdocs serve

Open http://localhost:8000.

Project layout¶

argus/
├── core/               # Pure Python — no cloud imports
│   ├── agent/          # ReAct loop, chat session, system prompt + tool schemas
│   ├── models/         # ResourceFinding dataclass
│   └── reports/        # Report builder, multi-cloud merge, export, notifications
├── adapters/
│   ├── base.py         # CloudAdapter abstract class
│   ├── aws/            # AWS adapter
│   ├── gcp/            # GCP adapter
│   └── azure/          # Azure adapter
├── ai/
│   ├── base.py         # AIProvider abstract class
│   ├── anthropic.py    # Anthropic direct API
│   ├── bedrock.py      # AWS Bedrock
│   ├── vertexai.py     # Vertex AI (Gemini)
│   └── azure_openai.py # Azure OpenAI (GPT-4o)
├── entrypoints/
│   ├── cli.py          # argus scan / argus chat subcommands
│   ├── cli_chat.py     # interactive chat REPL
│   ├── aws_lambda.py
│   ├── gcp_cloudrun.py
│   └── azure_function.py
├── deploy/
│   ├── aws/            # CloudFormation
│   ├── gcp/            # deploy.sh
│   └── azure/          # Bicep
├── examples/
│   └── chat_demo.py    # demo script for chat mode (no API key needed)
├── tests/              # mirrors source layout
└── docs/               # this documentation