We know that your source code is one of your most valuable and sensitive assets. For most development teams, it's a literal representation of your most important work product.
CodeStream’s servers do not require access to your source code as all interactions with the code are done locally in your IDE, leveraging your IDE’s access to the code. For example, when you select a block of code to discuss, CodeStream captures the commit IDs and line number offsets to allow us to maintain the connection between the discussion and the block of code.
Note that CodeStream does need to store snippets of code in our database. This happens when you discuss a block of code by creating a comment or issue. CodeStream saves that block of code in the database for a variety of reasons, not the least of which is the possibility that the code in question may not have been pushed and storing it would be the only way to share it with your teammates.