Exploit Prevention as a Service for Rails | Hacker News Reader

Exploit Prevention as a Service for Rails

hakiri.io

23 pointstkellogg13 comments

Open articleView on HN

Take my money! :) Your service is super-useful for app security. I'm signing up right now. Great idea.

My two cents for you: you're far enough along that you can start seeding your market as fast as possible.

* How about going to meetups at app dev companies such as Pivotal, Carbon 5, ThoughtWorks, etc.?

* How about a free version for open source projects, or students, or nonprofits?

Try to make the signup faster and easier.

* How about an item on the homepage saying "Upload a Gemfile.lock to see if it's secure"? I would personally do this first because it's fast, easy, and needs no setup.

* How about an item on the homepage asking "What's your GitHub username?" then skim for vulnerabilities? I would personally do this because I write many open source gems.

Great feedback!

A lot of these things are on the roadmap, including the free option for OSS, pitching at larger dev shops, and Gemfile.lock scanning without signing up.

Thank you.

This is a really useful service for what we call "version tracking," and it looks like it's running Brakeman for you too. With that said, it won't find vulnerabilities you code in yourself - only publicly released vulns that Ruby, etc. have issued patches for. A tool like https://www.tinfoilsecurity.com can help you find more vulnerabilities that either a) haven't been found yet publicly or b) you've written in yourself. (Disclosure: I'm the cofounder)

Definitely running Brakeman, as are https://codeclimate.com/, https://www.pullreview.com and http://rails-brakeman.com/.

Yup! Nothing wrong with that - just wanted to make sure it was clear what was going on. I run brakeman fairly often on my own for my projects. :)

"With that said, it won't find vulnerabilities you code in yourself - only publicly released vulns that Ruby, etc. have issued patches for."

This is not quite true. Brakeman (and Hakiri for that matter) runs static code analysis and finds vulnerabilities that were introduced by the developer. E.g., XSS, SQL injection, etc.

You're right - I took away too much credit by saying that. However, brakeman had its own issues (tons of false positives, etc.). It's a great project but too complex / cries wolf too often for a lot of projects. Dynamic analysis has issues too, but it often finds more than brakeman (or any purely static code analysis tool).

Love the idea. An upcoming project, that I am a part of, will likely be built on Rails, so this has been bookmarked!

Let me know if you need any help setting up :)

Automated security audits are useless until we have some fancy AI

Testing can only prove the presence of bugs, or in this case, exploits... but it can prove the presence of exploits, and that is valuable data. (At least, it can prove it reasonably well; false positives are a problem, but, well, that's life.) Absence of exploits may not prove you're safe, but presence certainly proves you aren't, and if your friendly pet automated scanner can find it, so can everybody else's automated scanner.

I'd be comfortable saying "this can give you a false sense of security"; useless is far too strong.

My guess is homakov was being sarcastic / kidding.

You're mostly right, but along with working scanner people should hire an auditor to fix real bugs. Or it is just useless, spending a lot of money annually for scanner forgetting about real vulns.

Reply on news.ycombinator.com