Each days I'm discovering open source project that pull my attention. Those projects are privacy friendly, game-changing or friendly to contribution. Here you can find a bunch of them listed by categories
OpenSSL is a robust, commercial-grade, full-featured Open Source Toolkit for the TLS (formerly SSL), DTLS and QUIC (currently client side only) protocols. The documentation isn't the best but you can use openSSL's library to include header like BigNum, AES, RSA and have fun with it.
Pycryptodome A self-contained cryptographic library for Python. I think this open source repository is the best for beginners to make their first contribution. Issues are easy to understand, the project isn't big compare to openSSL and you can have a good overview of the whole project.
MbedTLS An open source, portable, easy to use, readable and flexible TLS library, and reference implementation of the PSA Cryptography API. The documentation is way better than OpenSSL so you can more easily understand the library and code with it.
Zama-AI is an open source cryptography company building state-of-the-art FHE (Fully Homomorphic Encryption) solutions for blockchain and AI. The project is interesting because it introduces you to FHE and have a bounty program so if your are good and motivated you even can make a bit of money.
Cryptocoding is a repository maintained by Jean-Philippe Aumasson, a great cryptographer. It lists "coding rules" for implementations of cryptographic operations, and more generally for operations involving secret or sensitive values. This page is a very good ressource if you want to develop your own implementation. Of course those recommendations aren't enought for a well secured implementation but as a beginner it is a good training.
CS2P The Community Cryptography Specification Project is a project that facilitates the maintenance of cryptography specifications using software development methodologies. In other words, C2SP applies the successful processes of open source software development and maintenance to specification documents.
GUN is a peer2peer protocol that share data between several machines in live. With encryption, serializaion of the data and more you can easily build an application based on this protocol like alternative to Youtube, Zoom or Facebook. The protocol is implemented in JS and you can integrate it easily in your HTML with CDNs headers or directly in your JS code.
Stratosphere lab This amazing Czech laboratory publishes various of content such as dataset, tools, white papers all around the theme of detecting malicious connexion in a network. Their website is very wide and you have plenty of things to discover.
The Cyber Defense Matrix A framework created by @sounilyu to help you expertly navigate the cybersecurity landscape. Basically, this tool help you to visual what aspect of security you need to priorize for your structure or product. It has several use cases and the first dimension of the matrix is based on the NIST Cybersecurity Framework. Read the website to learn more.
Pigasus An IPS based on Snort fully implemented on an FPGA. With this method, the IPS can read a 100Gpbs network traffic without any loss.
Data Plane Development Kit This tool bypass the kernel to analyze network packages and make packet processing much faster.
Ollama is a LLM runners allowing you to interact locally with models like Lalla3, Mistral and more. You can interact with the model through a chat and you can use it for any of your needs. You don't even need a GPU, the minimal requirement is a recent CPU for decent performance (like 10th gen for Intel).
Open WebUI Open WebUI is an extensible, feature-rich, and user-friendly self-hosted WebUI designed to operate entirely offline. It supports various LLM runners, including Ollama and OpenAI-compatible APIs.
MLpack a fast, header-only C++ Machine Learning library. It aims to implement a wide array of machine learning methods and functions as a "swiss army knife
AB-TRAP is a step-by-step guide for building a ML model for IDS/IPS.
Real-CyberSecurity-Datasets Public datasets to help you tackle various cyber security problems using Machine Learning or other means.
Bend language A promising high-level, massively parallel programming language. Bend have the particularity to parallelize automatically every single mathematical operation allowing you to exploit 100% of your GPU without having to handle thread creation, locks, mutexes, or atomics. Bend is cross-platform, available on WSL2 and can support 10000 concurrent thread. Only Nvidia's GPU are supported for the moment.
Algorithmica is an open-access web book dedicated to the art and science of computing by Sergey Slotin. Its intended audience is everyone from performance engineers and practical algorithm researchers to undergraduate computer science students.
SimdJson Created by Daniel Lemire, SimdJson is tool parsing gigabytes of JSON per second. Used by Facebook/Meta Velox, the Node.js runtime and dozen of other real world usage, this tool is certainly one of the fastet in the world for json parsing.