Caprock Files

Added

• Pick a folder instead of typing its path. Starting a session meant typing an absolute path from memory into a dashboard that is already showing the repositories you work in every day.

Two lists, in the order they are useful. Recent is where sessions have already run, newest first — for most people that is the whole picker, since the next session is almost always in a repository you were in yesterday. Browse walks down from one root for the first session in a new project; repositories are marked and sorted first, because a repository is what you are looking for and everything else is the route to it.

The text field stays: it is the fastest input for anyone who knows the path, it is what a paste goes into, and the lists write into it rather than replacing it, so what will actually be used stays visible and editable.

The root is yours to set — browse_root, defaulting to your home directory. It is a setting because "where I keep my code" is personal, and because the narrower it is, the less the daemon's directory listing can be asked for.

Browsing is the only thing in Caprock that reads the filesystem for a web page, so the boundary is the feature: directories and names only, never contents; symlinks resolved before the containment check, so a link inside the root pointing out of it does not slip through; dotfile directories never listed; and "outside the root" and "does not exist" answering with the identical 404, so it cannot be used to test whether a path exists. Each of those is covered by a test that was verified by breaking it.