Jump to content

toor

Members
  • Content Count

    2
  • Joined

  • Last visited

  • Days Won

    1

Everything posted by toor

  1. Is it possible to lock a browser extension to specific vaults only? I use multiple browsers / chrome profiles and multiple vaults. It would be great if there was an option to lock these to only be able to read specific vaults. For example, I don't want my work chrome profile or browser reading from my non-work vault and vice versa.
  2. toor

    Security audit

    I've been following Enpass for a while but have never seen a need to comment on the forum since I was waiting for a security audit before purchasing. I work in this area and I want to clarify a few things on here: First of all, the disclaimer "It is important to note that because of the time constrains naturally involved during a Penetration Test exercise this project should not be considered a full security audit", is standard. You're unlikely to going to find someone who is going to declare something secure and take ownership of any vulnerabilities that are found. By their nature any audits are going to be limited in time and have disclaimers. A two week audit by two people is quite expensive but is still best effort. Windows was audited for years by a multitude of people before being released, yet they still had a bunch of vulnerabilities. That being said, from my experience a two person two-week audit is probably enough for a smaller project like this if you exclude the open source software that it uses - and given the concerns people have being due to the software being closed source, that's probably fair. There's no point in spending two weeks auditing SQLCipher when people are worried about Enpass itself. Now I do have some concerns with respect to the audit. There seems to be very little information about what they tested - if anything - other than trying to extract the master password in a variety of ways. Did they look for potential memory corruption vulnerabilities? Did they test the "password sharing" feature that is new and is an obvious point of attack. Did they test the browser plugins, which are another possible attack vector? They mention looking at restoring databases, that's definitely an area of attack: say you store a less important database in the cloud, could it be used to compromise the application when it opens this database (possibly this vecotr only affects SQLCipher so it may have been out of scope)? Did they consider these attack vectors or were they only looking for master password issues? From their summary and methodology it seems that they would have, but there is too little information on this. Another concern that I have with the audit is the following: How much time was wasted reverse engineering Enpass v 5.6.9 before the source code was provided for 6? This is less of a concern for Android since Java applications are easily reversible, but they were still looking at older code at the time. How quickly did they get access to the Windows source code? There's a big difference between a one-week source code assessment and a two-week source code assessment. Someone mentioned PCI on this forum, that is only done for payment processing (you can tell by the name Payment Card Industry Data Security Standard). As far as I can tell Enpass does not take payments, they only allow purchases via app stores, thus have no need for PCI. In general PCI is a checklist for minimum standards: do you have a firewall, do you encrypt payment card data at rest and in transmission, etc. That checklist is then verified by an auditor, but it's meant to satisfy the payment processors and says nothing about the security of the software that Sinew produces. That being said, I want to applaud Enpass for making the full report accessible, very few companies would provide the report to their customers in full and would simply say "we've been audited by X".
×
×
  • Create New...