Blog entry, Stardate April, 2005

We have recently been on a new shellcode kick, but this time it's not about
making them smaller. We're currently working on building very power new
post-exploitation shellcode and toolkits, and a very powerful unified API
to expose their functionality. This will allow us to diverge from precanned
payloads, allowing users to quickly build power and portable post-exploitation
tools. We've have built strongly upon our Windows DLL injection, and are
working on extending similar functionality to the land of Unix.  We're also
working hard on the next version of Metasploit, which follows this same
philosophy of emedability and extensablity, allowing users to build their own
tools on top of our framework. Our previous work was all about exploit
frameworks. Our new approach we are really building more of a hacker tool
framework, allowing very strong automation and customization.



ignore this.....

With Metasploit 3.0 on the horizon, we've been working hard on design and
building components to take a very different focus.  Currently Metasploit 2
is very much an end user tool, and doing anything custom isn't so elegant.
The approach we are taking in Metasploit 3, is "Metasploit as a library".  We
are working very hard on writing post-exploitation suites for different
platforms, and then building a unified API that they all adhere to.  This will
allow you to directly script remote hosts, proxying file operations, network
communications, and transparent channelized communication.  The new system is
being designed to be threadsafe to a high degree, forcing us to rethink and
redesign many of our tools and protocols.  Our new system is design to allow
you to do things like run 10 exploits concurrently, all "pivoting" through
a host you previously owned.  Along with this new feature set, we are still
concentrating on keeping things off disk, all in-memory injection, etc.