CONTRIBUTING.md

Contributing

Thanks for contributing to the Confidential ML Feature Store.

Development workflow

1. Create a feature branch from main.
2. Keep changes scoped and reviewable.
3. Run formatting, linting, type checks, and tests before opening a pull request.
4. Update documentation when behavior, configuration, scripts, or endpoints change.

Local setup

Install dependencies:

make install

Common commands:

make format
make lint
make test
make coverage

For local mock-enclave development:

docker compose up -d dynamodb-local
make run-dev-mock

Code standards

• Preserve the current file structure and public interfaces unless there is a strong reason to change them.
• Use type hints throughout.
• Add or update docstrings for public functions, classes, and methods.
• Prefer explicit, meaningful error messages over generic failures.
• Keep request/response schemas and routes aligned with the existing FastAPI application.

Tests

This project uses:

• pytest
• moto
• local mock components such as MockEnclaveClient and MockNSM

Before submitting a PR, make sure all tests pass:

make test

If you changed typing or formatting-sensitive code, also run:

make lint

Documentation

When changing behavior in any of these areas, update the docs in the same PR:

• endpoint routes or schemas
• environment variables
• AWS setup steps
• scripts in scripts/
• enclave / mock-enclave behavior

If you add or update screenshots, place them in docs/screenshots/ and update docs/screenshots/README.md.

Pull requests

A good pull request should include:

• a concise summary of the change
• the reason for the change
• any migration or setup notes
• test evidence when behavior changed