Skip to main content

Compliance testing for implementations of cryptographic primitives

Project description

crypto-condor

The logo of crypto-condor, depicting a condor holding a key.

crypto-condor is a tool for compliance testing of cryptographic primitives, in the form of a Python library and CLI. It is complemented by an extensive documentation, featuring guides on the primitives supported.

Q: What is a cryptographic primitive?

A: A low-level cryptographic algorithm, generally used to build a protocol. For example, AES is an encryption primitive that is used in the TLS protocol, which is the protocol your browser used to securely get this page.

Q: What is compliance testing?

A: Algorithms are described in specifications, such as FIPS publications or RFCs. When implementing these algorithms, we want to ensure that they comply with the specification, i.e. the implementation behaves as the algorithm described.

Q: How to test for compliance then?

A: We can use test vectors, which are sets of inputs and their corresponding outputs. For example, encrypting with AES is a deterministic operation: for a given key and message, AES will always return the same ciphertext. So we can choose some input values, run the algorithm, and record the value returned. All implementations of AES are then expected to return the same ciphertext for this given key and message. If it does not, then it is not compliant.

Q: And so, what does crypto-condor do?

A: crypto-condor provides both a nice Python API and a wrapper system to test implementations with sets of test vectors that come from sources such as the NIST CAVP.

The Python API exposes test functions that take an implementation as input, in the form of a Python function or class, passes the inputs defined by the test vectors to that implementation, and checks if the outputs are those defined by the vectors.

The wrappers are small programs that already define the function prototype. The user calls the implementation to test inside this function, and crypto-condor runs it with the test vectors as with the Python API.

And it comes with a documentation, wrapper examples, and guides on supported primitives.

Requirements

crypto-condor requires Python 3.10+. For information, it is developed using Python 3.12 on Fedora 39.

The implementations of AES, Kyber, Dilithium, and TestU01 are written in C and are compiled directly on the user's machine. As such, they require a C compiler and GNU Make. These primitives are only compiled when required, and not when installing the package:

  • AES: when testing the output of an implementation using classic modes of operation (not CCM or GCM).
  • Kyber: when testing the output of an implementation or when using test vectors on the encapsulate function.
  • Dilithium: when testing the output of an implementation.
  • TestU01: when used to test a file.

Installation

It is available on PyPI:

python -m pip install crypto-condor

An up-to-date list of the requirements can be found in the [tool.poetry.dependencies] section of the pyproject.toml file.

Usage

The documentation is available at https://quarkslab.github.io/crypto-condor/latest/index.html.

Once installed, the CLI is available as crypto-condor-cli. It is structured in commands, similar to Git. Run it without arguments or with --help to display the help message detailing the available subcommands. You can check the documentation for a quick rundown of all the commands.

As for the Python library, it is available as crypto_condor (note the underscore). Each primitive has its own module under primitives, e.g. crypto_condor.primitives.AES. It contains the functions used to test implementations.

Development

See CONTRIBUTING.

Changelog and versioning

A changelog is available. This projects adheres to CalVer. The format used is YYYY.MM.DD[.MICRO][-MODIFIER]:

  • YYYY: full year (2023).
  • 0M: zero-padded month (01, 02, ..., 12).
  • 0D: zero-padded day (01, 02, ..., 31).
  • MICRO: an increasing counter, used for patches published in the same day.
  • MODIFIER: usually rc<n> to indicate a release candidate.

Authors

  • Julio Loayza Meneses, Quarkslab.
  • Angèle Bossuat, Quarkslab.
  • Dahmun Goudarzi, Quarkslab.

Logo idea by Robin David, drawing by Irene Loayza.

License

Apache License Version 2.0, see LICENSE.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

crypto_condor-2024.9.24.tar.gz (75.9 MB view details)

Uploaded Source

Built Distribution

crypto_condor-2024.9.24-py3-none-any.whl (76.3 MB view details)

Uploaded Python 3

File details

Details for the file crypto_condor-2024.9.24.tar.gz.

File metadata

  • Download URL: crypto_condor-2024.9.24.tar.gz
  • Upload date:
  • Size: 75.9 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/5.1.1 CPython/3.12.6

File hashes

Hashes for crypto_condor-2024.9.24.tar.gz
Algorithm Hash digest
SHA256 a15ae7bb82a5f9053074ab9551255ca11e15058ecbcc3a4ba0a0738e9c01a8ab
MD5 37ef419d1a34da565d42d8f2dc819d3d
BLAKE2b-256 d3b5ab8ca25dc2b6c722899c57f1969f63882743c33b024cf7e489ccfcf79196

See more details on using hashes here.

File details

Details for the file crypto_condor-2024.9.24-py3-none-any.whl.

File metadata

File hashes

Hashes for crypto_condor-2024.9.24-py3-none-any.whl
Algorithm Hash digest
SHA256 bbd2cfdd728386321cdec68cd334fcee3f4ae24b63d265db0ab55bce49a103ac
MD5 920cc258c931e6a151075091cdadb015
BLAKE2b-256 1dd075f4c924dc5714dca79bd3906deaf423fc2985b71c54d2ad1e1126ba2ed3

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page