kron — Modern Cron Replacement

kron is a CLI-first, single-binary cron replacement written in Rust. It solves cron's observability black hole: every run is automatically captured with stdout, stderr, exit code, and duration. No silent failures.

When to Use This Skill

• User wants to schedule a recurring job or task
• User wants to check if a scheduled job ran successfully
• User wants to see output/logs from a past job run
• User mentions cron, crontab, job scheduling, or kron
• User wants to migrate from system crontab

Installation

cargo install kron

Single binary, zero runtime dependencies.

Core Commands

Add a job

# With a cron expression
kron add --name backup "0 2 * * *" pg_dump mydb -f /backups/mydb.sql

# With human-readable schedule
kron add --name backup "every day at 2am" pg_dump mydb -f /backups/mydb.sql

# Every minute (for testing)
kron add --name heartbeat "* * * * *" echo "alive"

# Capture current environment variables
kron add --capture-env --name deploy "0 3 * * *" ./deploy.sh

The --name flag is optional. The schedule comes first (quoted cron expression or human-readable), then the command to run.

List all jobs

kron list

Output:

ID       NAME                 ENABLED  SCHEDULE       RUNS   COMMAND
──────────────────────────────────────────────────────────────────────
a1b2c3d4 backup               yes      0 2 * * *      5/5    pg_dump mydb -f /backups/mydb.sql
e5f6a7b8 heartbeat            yes      * * * * *      10/10  echo "alive"

Jobs can be referenced by ID (or prefix) or name in all commands.

Import from crontab

kron import              # interactive selection
kron import --all        # import all entries

Reads the current user's crontab, displays entries in a numbered list, and lets you select which to import (e.g. 1,3-5 or all). Handles @hourly/@daily/etc. aliases, skips comments, env vars, and @reboot entries. Derives job names from inline comments or command basenames.

Force-run a job now

kron run backup

Executes the job immediately (outside the schedule), captures output, and records the run.

Dry-run a job

kron test backup

Executes the job but does not record the run. Useful for testing commands.

Check job status

kron status

Output:

NAME                 STATUS     EXIT CODE    LAST RUN
-----------------------------------------------------------------
backup               success    0            2026-03-15 02:00:01 (3s)
heartbeat            success    0            2026-03-15 15:59:00 (0s)

All timestamps are displayed in local timezone.

View run history

kron history backup
kron history backup -n 20   # last 20 runs
kron history                 # shows all jobs (most recent first)

Output:

#    STATUS     EXIT CODE    DURATION   STARTED
-----------------------------------------------------------------
1    success    0            3s         2026-03-15 02:00:01
2    success    0            3s         2026-03-14 02:00:01
3    failed     1            1s         2026-03-13 02:00:00

View logs from a run

kron logs backup           # most recent run
kron logs backup --run 3   # specific run
kron logs                  # most recent run across all jobs

Output:

=== Job: backup | Run #1 | 2026-03-15 02:00:01 ===
Status: success | Exit code: 0
Duration: 3s

--- stdout ---
DUMP DATABASE mydb
...

Remove a job

kron remove backup

Manage alerts

kron alert add-telegram --token <BOT_TOKEN> --chat-id <CHAT_ID>
kron alert add-slack --webhook-url <URL>
kron alert add-webhook --url <URL>
kron alert list
kron alert test            # send test notification to all providers
kron alert remove <index>

Daemon management

kron daemon start              # run scheduler in background
kron daemon start --foreground # explicit foreground mode
kron daemon stop               # stop the running daemon
kron daemon install            # install as systemd/launchd service
kron daemon uninstall          # remove system service
kron daemon status             # check service status

Self-update

kron update

Job Configuration (TOML)

Jobs are stored as TOML files in ~/.config/kron/jobs/<id>.toml:

[job]
id = "a1b2c3d4"
name = "backup"
command = "pg_dump mydb -f /backups/mydb.sql"
schedule = "0 2 * * *"
working_dir = "/app"
enabled = true
timeout = "30m"

[job.env]
DATABASE_URL = "postgres://localhost/mydb"

[job.alert]
on_failure = true
on_success = false
on_silence = "1h"

Cron Expression Reference

┌───────────── minute (0-59)
│ ┌───────────── hour (0-23)
│ │ ┌───────────── day of month (1-31)
│ │ │ ┌───────────── month (1-12)
│ │ │ │ ┌───────────── day of week (0-6, Sun=0)
│ │ │ │ │
* * * * *

Common patterns:

Data Storage

• Job definitions: ~/.config/kron/jobs/*.toml (single source of truth)
• Run history: ~/.local/share/kron/kron.db (SQLite, WAL mode)
• Alert config: ~/.config/kron/alerts.toml

Global Configuration

Optional global settings in ~/.config/kron/config.toml:

[retention]
max_runs_per_job = 100   # default: 100
max_age_days = 30        # default: 30

The daemon automatically prunes old runs based on this policy.

Built-in Safety Features

• Overlap prevention: A job won't start if a previous instance is still running
• Timeout: Jobs can be killed after a configurable duration (kill_on_drop — no orphan processes)
• Output capture: stdout and stderr always recorded (max 1 MiB each), never lost
• Graceful shutdown: SIGINT/SIGTERM cleanly stops the daemon
• SIGHUP reload: kron add/kron remove signal the daemon for instant config pickup
• Duplicate prevention: Daemon refuses to start if another instance is already running
• Automatic pruning: Old runs cleaned up based on retention policy (default: 100 runs/job, 30 days)
• File permissions: Job files 0600, jobs dir 0700, DB file 0600, alerts.toml 0600
• Next-fire scheduling: Zero idle wakeups — daemon sleeps until the next job is due

Common Workflows

Migrate from crontab

kron import                  # select entries interactively
kron daemon install          # install as system service
crontab -r                   # remove old crontab (after verifying)

Debug a failing job

kron status              # See which jobs are failing
kron history myjob       # See recent run history
kron logs myjob          # See the actual error output
kron test myjob          # Dry-run to test without recording
kron run myjob           # Re-run and record the result

Disable a job temporarily

Edit ~/.config/kron/jobs/<id>.toml and set enabled = false. The daemon picks up manual edits within 60 seconds, or send kill -HUP $(cat ~/.local/share/kron/daemon.pid) for instant reload.