Point Cronhq at a webhook. We fire it on schedule, retry on failure, and page you when something breaks — and again the moment it recovers.
curl -X POST https://api.cronhq.xyz/v1/jobs \
-H "Authorization: Bearer chq_••••" \
-d '{
"name": "nightly-rollup",
"schedule": "0 2 * * *",
"webhook_url": "https://api.you.com/cron/rollup"
}'Throughput
312/min
P95 latency
187ms
Success
99.94%
↳ exec feed · mock data, real shape
// shipped this cycle — the stuff that keeps you here
01 · Signed webhooks
timestamp.body
│
hmac-sha256 ── secret
│
▼
v1=20d7b6d0e4f1…Every request carries an X-Cronhq-Signature header — HMAC-SHA256 over timestamp.body — plus an X-Cronhq-Timestamp. Each job gets its own secret. Rotate it anytime, no downtime.
verify webhooks →02 · Heartbeat monitors
ping ✓ ✓ ✓ ✕
└─ DOWN
window missed ─┘Ping a URL on every run. Miss the window — period plus grace — and we flip the monitor DOWN and page you once. It's cron's inverse: proof of absence, not presence.
add a monitor →03 · Cron-as-code
$ npx cronhq jobs tail nightly-rollup
success200412ms
npx cronhq sync reconciles a cronhq.yaml in your repo — creates, updates, prunes. npx cronhq tail live-streams executions straight to your terminal.
read the CLI docs →04 · ⌘K + natural language
"every weekday at 9am"
│
▼
0 9 * * 1-5Hit ⌘K anywhere in the dashboard for the command palette. Type "every weekday at 9am" and get a valid cron expression back, with a plain-English preview so you never guess wrong.
try the dashboard →Sign up with your email and we'll send a one-time link — click it and a key is minted for you: a 36-character string starting with chq_. Already have one? Sign in with what you've got.
Pick a schedule (every 5 min, daily at 9am, weekdays only…) and the URL we should call. One request, two seconds, done.
POST /v1/jobs
{
"name": "nightly-rollup",
"schedule": "0 2 * * *",
"webhook_url": "https://api.you.com/cron/rollup",
"timezone": "America/New_York"
}Every execution is logged with status, duration, HTTP code, and the response body. Newest first. Searchable. Yours forever.
Wire email or Slack. We page on the third failure in an hour (not every failure), and again the moment things recover.
nightly-rollup recovered after 3 failures. Last 6 runs: all 2xx.
01 · Exactly once
┌─────────┐ ┌─────────┐
│ worker A │ │ worker B │
└────┬────┘ └────┬────┘
└──── LOCK ────┘
▼
┌─────────┐
│ job 0x4 │
└─────────┘Two workers can never fire the same scheduled execution. Crashed mid-job? Another picks up after the lock expires.
02 · Retries built in
try ✕──┐
+60s ✕──┤
+120s ✕──┤
+240s ✓──┘
▼
okMax retries + delay per job. The terminal status and last error always land in your history — so the postmortem writes itself.
03 · Alerts with dedup
✕ ✕ ✕ ━━ alert
✕ ✕ ✕ (silent)
✕ ✕ ✓ ━━ recovery
(silent)Failure alert fires on the third bad run in an hour, not the first. Recovery alert fires on the first success after a streak.