Data Storage

All Hourgit data is stored locally on your machine. There are no servers or cloud sync.

File Locations

Path	Purpose
`~/.hourgit/config.json`	Global config — defaults, projects (id, name, slug, repos, schedules)
`REPO/.git/.hourgit`	Per-repo project assignment (project name + project ID)
`~/.hourgit/<slug>/<hash>`	Per-project entries (one JSON file per entry)
`~/.hourgit/watch.pid`	PID file for the filesystem watcher daemon (precise mode)
`~/.hourgit/watch.state`	Watcher state file — last activity timestamps per repo (precise mode)

Entry Types

Each entry is a JSON file identified by a 7-character hex hash (similar to git commit hashes). The type field distinguishes between:

log — manually logged time entry (duration, start time, message, task label)
checkout — branch checkout event recorded by the git hook (previous branch, next branch, timestamp, repo)
commit — git commit event from reflog (commit ref, timestamp, message, branch, repo); used to split checkout sessions into finer time blocks
submit — submission marker for a report period (date range, creation timestamp)
activity_stop — idle detection: records when file activity stops (timestamp of last file change, repo path)
activity_start — idle detection: records when file activity resumes (timestamp, repo path)

Projects

Projects are defined in ~/.hourgit/config.json and contain:

id — unique identifier
name — display name
slug — filesystem-safe name (used as directory name under ~/.hourgit/)
repos — list of assigned repository paths
schedules — per-project working hours configuration

Per-Repo Assignment

When you run hourgit init or hourgit project assign in a git repository, a .hourgit file is created inside the repo's .git/ directory. This file maps the repository to a project without modifying tracked files.