Reference
The lookup tables the usage guide points at. A cron folder is plain JavaScript; the only magic is the entry-point filenames. These pages spell out that grammar, the module contract a script fulfils, the commands that drive the client and server, and every file that can live in a content root — so each guide page can link to one source of truth.
Schedules →
The definitive filename grammar: every interval unit, every clock token, and the valid range of each — plus base-name extraction and the mutex.
Scripts →
The script module contract: what a cron script may export, the ctx it receives,
success and failure semantics, and how require resolves.
Commands →
The client CLI flags and the server-side service commands — starting, restarting and inspecting the API and runner, and pulling logs back.
Files →
The content-root layout: naming rules, the ignore rules, supporting files, and the
_env.json secrets file — each path shown as a pattern.
The shape of a content root
One flat script directory, mirrored whole. Top-level schedule-named files are jobs; everything else rides along as supporting code.
cron/ your content root, mirrored as a whole
digest.every.30m.cron.js entry point → runs every 30 minutes
backup.daily.0300.cron.js entry point → runs daily at 03:00 UTC
report.hourly.00.cron.js entry point → at :00 past every hour
report.hourly.30.cron.js entry point → same base name, at :30 too
lib/ supporting modules, imported by jobs
format.js
mailer.js
_draft.every.10s.cron.js ignored (leading _ = local only)
_env.json ignored secrets, read by the runner
.gitignore ignored (dotfile)
The Files page breaks this down path by path; the Schedules page is the grammar behind the entry-point names.