Files and settings
| What | Where |
|---|---|
| Config | ~/Library/Application Support/Wake/config.json |
Listener state (wake status reads it) |
~/Library/Application Support/Wake/state.json |
Your runs, by name (wake open reads it; the last 200) |
~/Library/Application Support/Wake/runs.json |
Worktrees Wake made (wake prune removes only these) |
~/Library/Application Support/Wake/worktrees.json |
| Repositories each app has asked for (setup offers them first) | ~/Library/Application Support/Wake/asked-repos.json |
| Log | ~/Library/Logs/Wake/wake.log |
| Place keys | The macOS Keychain, service dev.abyssinia.wake |
| LaunchAgent | ~/Library/LaunchAgents/dev.abyssinia.wake.plist |
| Claude Code settings for herdr runs | ~/.wake/runs/<session>.settings.json |
The config, state, run history and log are readable by you alone (0600, in 0700
folders). Every setting has a command (see Commands);
editing the file by hand works too, and the listener reads it within half a
minute. bypassPermissions in the file is read as the default.
Set WAKE_HOME to keep everything in another folder, for testing.
The LaunchAgent is written with the PATH of the shell you ran
wake install from, so it finds your agents, git and gh. Run it
again after you install a new agent somewhere else on your PATH.