Open source software is in our company's DNA.


Alejandro and Amanda, our co-founders, started Hyperion Gray in 2012 with the launch of our flagship open source project called Punkspider, a distributed, internet-scale web vulnerability fuzzer. That project led to our first DARPA grant, and we've been on this wild ride of technology R&D ever since.

In the course of our R&D work we've built a lot of things and solved a lot of problems, and we've open sourced all sorts of scripts, tools, utilities, libraries and applications. Beyond our funded R&D, we also occasionally release random side projects, and we even write poetry! All of our projects are built on open source stacks—standing on the shoulders of giants—and giving back is a core part of our mission. Sure, it's tough to build a business model around giving almost everything away for free, but we think it's best way to achieve our goal of making a positive impact in the world.

Featured Projects

These are some of our favorite open source projects that we released.
  • Agnostic Database Migrations: A light-weight, easy-to-learn, and flexible database migration tool in which migration scripts are written in pure SQL. It is agnostic towards database, programming language, and object relational mapper (ORM).
    Documentation Source Code MIT License
  • Aquarium: A clustered configuration for the Splash headless browser to permit greater throughput and also better fault-tolerance than stand-alone Splash.
    Source Code MIT License
  • BK String: A fast implementation of Burkhard-Keller trees written in Rust with a Python wrapper.
    Source Code MIT License
  • Dropchat: A clientless, safe, disposable chat service for safe and anonymous communication. Operates over an ephemeral Tor hidden service.
    Source Code MIT License
  • ELI5: Explain it like I'm 5: a framework for visualizing and debugging black box machine learning algorithms such as decision trees, boosting models, and conditional random field (CRF) classifiers. Also implements LIME technique for explaining arbitarily complex models such as neural networks.
    Documentation Source Code MIT License
  • Formasaurus: A machine learning system that analyzes HTML documents and automatically classifies form elements into categories such as login forms, search forms, etc.
    Documentation Source Code MIT License
  • Google Spider: Extract result URLs from Google search queries. (This project is for educational puposes only.)
    Source Code MIT License
  • MaybeDont: A crawling library that uses machine learning model to predict which URLs will lead to duplicate content. Avoiding duplicate content improves crawl efficiency.
    Source Code MIT License
  • Soft404: A classifier for detecting soft 404 pages. A "soft 404" is a page that is served with 200 status but has human-readable content indicating that the content is not found.
    Source Code Apache License
  • Site Hound: Site Hound is a domain discovery Tool that extends the capabilities of commercial search engines using automation and human-in-the-loop (HITL) machine learning, allowing the user efficiently expand the set of web pages relevant to a topic of interest.
    Source Code Apache License
  • Starbelly: An agile, user-friendly, streaming web crawler. All crawling behavior is configurable through a graphical interface. Uses machine learning to simplify common tasks like logging in.
    Documentation Source Code MIT License
This is just a small sample of our open source projects. For a comprehensive listing, take a look at our corporate GitHub and Memex GitHub accounts, as well as the official Memex Program Index.

Online Courses

We also created an open source course: How to Hack and Defend Your Website in 3 Hours. It's perfect for beginners, so take the course and then try your hand at the Capture The Flag!