Task Scheduling

Five CLI commands drive Webrium's task scheduler:

• make:schedule — scaffold a new task file in app/Schedules
• schedule:run — run every task that is due right now, then exit (what a real cron entry calls)
• schedule:list — list every registered task with its next due time
• schedule:test — run a single task immediately, ignoring its schedule
• schedule:work — run the scheduler in the foreground for local development, without a system cron entry

Defining tasks themselves — Schedule::call(), the frequency methods, naming, file discovery — is documented in Core → Task Scheduling. This page focuses on the CLI surface and deployment.

make:schedule

php webrium make:schedule <name> [--force]

Creates app/Schedules/<name>.php from a template. Creates the app/Schedules directory first if it doesn't exist yet — useful when upgrading a project created before the scheduler existed.

Examples

# Create app/Schedules/SendReports.php
php webrium make:schedule SendReports

# Overwrite an existing file
php webrium make:schedule SendReports --force

schedule:run

php webrium schedule:run

Loads every task file under app/Schedules and runs whichever tasks are due at this exact moment, then exits. This is the command a real system cron entry should call, once a minute:

* * * * * cd /path/to/your/project && php webrium schedule:run >> /dev/null 2>&1

Webrium itself decides which tasks are actually due each time this fires — you never need a separate crontab line per task.

Output & Exit Code

Each due task prints one line reporting whether it ran, was skipped (its lock was already held), or failed:

✔ reports.daily
⚠ cleanup.temp skipped (already running)
✘ newsletter.weekly failed: Connection timed out

A load error (a broken task file) is reported the same way, before the task list:

✘ Failed to load /path/to/app/Schedules/Broken.php: syntax error, unexpected...

Nothing here stops the batch. A broken file doesn't prevent other files from loading, and a failed task doesn't prevent other due tasks from running — see Core → Task Scheduling → Running Tasks for why. The command's exit code does reflect whether anything went wrong (0 if every load and every due task succeeded, 1 otherwise), so cron's own failure notifications (mail-on-error, monitoring) still work — this is checked only after everything that was going to run has run.

schedule:list

php webrium schedule:list

Lists every registered task with its cron expression and next due time — read-only, runs nothing:

+---------------+-------------+------------------+
| Task          | Expression  | Next Due         |
+---------------+-------------+------------------+
| cleanup.temp  | */5 * * * * | 2026-09-19 12:05 |
| reports.daily | 0 1 * * *   | 2026-09-20 01:00 |
+---------------+-------------+------------------+

Useful for confirming a new task registered correctly, and for sanity-checking when it will actually next fire before you go looking for it in a log.

schedule:test

php webrium schedule:test [<name>]

Runs one task immediately, regardless of its schedule — for verifying a task works without waiting for (or temporarily changing) its due time. Still goes through the same overlap lock and error isolation as schedule:run.

Argument

Examples

# Run a specific task by name
php webrium schedule:test reports.daily

# No name given — pick one interactively
php webrium schedule:test

Which scheduled task would you like to run?:
  [0] cleanup.temp
  [1] reports.daily
 > 1
Running 'reports.daily'...
✔ reports.daily

schedule:work

php webrium schedule:work

Runs the scheduler in the foreground: wakes up at the start of every minute, reloads app/Schedules from scratch, and runs whatever is due — the same load/run isolation as schedule:run, just looped, so you don't need a real cron entry configured on your development machine.

Stop it with Ctrl+C (or SIGTERM). Where the pcntl PHP extension is available, this is graceful — the current tick finishes before the process exits and prints Schedule worker stopped.; without pcntl the process simply ends immediately, like any other command would.

Not a production substitute for real cron. schedule:work is one long-lived process: if it crashes or the server reboots, nothing restarts it on its own, unlike a cron entry (which the OS re-evaluates independently every minute regardless of what happened the minute before). Use schedule:run behind real cron in production; reach for schedule:work for local development only.

A Typical Workflow

Local development — no crontab needed:

php webrium make:schedule SendReports
# ... write the task, using Schedule::call() as documented in Core → Task Scheduling ...
php webrium schedule:list             # confirm it registered, check next due time
php webrium schedule:test reports.daily  # run it once, right now, to verify it works
php webrium schedule:work             # leave it running while you iterate

Production — one cron entry, forever:

* * * * * cd /var/www/your-app && php webrium schedule:run >> /dev/null 2>&1

New tasks added later don't need a new crontab line — schedule:run already picks up anything registered under app/Schedules on its next run.