Triggers & Scheduling

Run when it matters.
Manual, scheduled, webhook, conditional.

Cron schedules with timezone-aware precision. Webhook endpoints with HMAC signature verification. Manual one-off runs. Conditional triggers fired by upstream data. Pause for human approval at any step, resume by Slack reaction or email click. Every run logged, every input replayable.

Get a live workflow build → Back to Agents

app.demandsphere.com - Agent / Schedule

Trigger type

Scheduled

0 9 * * MON

Webhook

HMAC signed

Conditional

if rank drop > 5

Manual

run now

Fired

Mon 9:00am

Data

Fetch ranks

LLM

Summarize

Paused - approval

Publish 3 fixes?

ApproveReject

Deliver

Slack

Waiting on approval from #seo-leads Auto-approve in 3h

Trigger features

Scheduling that doesn't pretend production is easy.

Cron with timezone awareness

Minute-level precision. Common presets (hourly, daily, weekly, monthly) in the UI; custom cron for anything else. Schedules respect the agent's timezone, not the server's.

HMAC-signed webhooks

Every webhook endpoint gets a per-agent secret. Incoming requests must sign the body with HMAC-SHA256. Mismatched signatures reject before any node runs.

Conditional triggers

Fire the workflow only when upstream data matches. Example: run a citation-recovery agent only if AI Overview presence drops by >10% in a tracked query set.

Human-in-the-loop pause

Any node can pause for review. Approval lands in Slack or email. One-click resume. Auto-approve timers if no response in N hours. Audit trail on every decision.

Replay any run

Every run logged with full input, node-level execution trace, and output. Replay historical runs with same or modified input. Useful for debugging, regression testing, or recovering from downstream failure.

Retry & backoff

Per-node retry policy with exponential backoff. Fallback paths for transient LLM provider errors. Idle states don't burn credits.

Secure & on schedule

Verified webhooks. Predictable runs.

Incoming webhook signature valid

POST /trigger/abc123
content-type: application/json
x-agent-signature: sha256=9f86d08...
// platform recomputes HMAC over body
✓ match -> run starts
✗ mismatch -> 401, no node runs

Next scheduled runs

Weekly Citation ReportMon 09:00

Daily AI Overview checkDaily 07:00

Rank-drop watcherconditional

Monthly Exec Digest1st 08:00

What runs after the trigger

Visual Builder →

The canvas where triggers connect to the rest of the workflow.

LLM Analysis →

What you commonly run after a scheduled or webhook trigger.

Connections →

Where the output lands - Slack, email, your webhook, your CMS.

Templates →

Scheduled-trigger templates: weekly reports, daily audits, monthly digests.

Common questions

Frequently asked

Manual (run now), Scheduled (cron expressions, timezone-aware), Webhook (HTTPS endpoint with HMAC signature verification), and Conditional (run when an upstream data node returns matching results).

Cron-style scheduling with minute-level precision, timezone aware. Common patterns (hourly, daily 9am, weekly Mon, monthly first) are presets in the UI. Custom cron expressions are supported for anything more specific.

Every webhook endpoint gets a unique HMAC secret. Incoming requests must include a valid X-Agent-Signature header computed from the secret and the request body. Failed verification rejects the run before any node executes.

Yes. Every run is logged with full input, output, and node-level execution trace. Replay any historical run with the same input or modified input; useful for debugging, regression testing, or recovering from a downstream failure.

Trigger your first agent

Book a strategy session and we'll wire up a scheduled or webhook-triggered agent on your data.

Get a live workflow build → See Pricing

Run when it matters.Manual, scheduled, webhook, conditional.