New Orchly MCP & API are live. Build custom workflows with your SEO & AI visibility data.

How to schedule and trigger agents in Orchly

Run agents once, on a recurring schedule, from a webhook, or by chat. Learn how to set the right trigger for each job.

Updated on June 13, 2026

An agent is only useful if it runs at the right time. Some jobs you kick off by hand; others should run on their own every week without you thinking about it. Orchly lets you do both, and choose whether the agent finishes the job or hands it to you for review.

Run it by hand

The simplest trigger is manual: you open the agent and run it when you want. This is how you test a new agent, and it’s right for one-off jobs like translating a single page or drafting one article.

Run it on a schedule

For recurring work, give the agent a schedule so it runs on its own. Set how often it should run (for example, every week) and Orchly handles it from there. Scheduling is what turns an agent from a tool you use into a teammate who just does the work:

  • A refresher agent that updates your oldest pages every month.
  • A monitoring agent that checks visibility every Monday and posts a summary to Slack.
  • A writer that drafts a new article each week against your content plan.
Stagger your schedules

If you run several agents, spread their schedules across the week instead of all at once. You’ll get a steady stream of work to review rather than a Monday-morning pile, and it’s easier to keep quality high.

Autopilot or review

Separate from when it runs is what happens when it finishes. Agents work in two modes:

  • Review, the agent does the work and queues it for you to approve before anything goes live. Safest, and how the prebuilt Article Writer behaves.
  • Autopilot, the agent completes the job and publishes on its own, no approval step.
Earn your way to autopilot

Run a new scheduled agent in review mode for a few cycles first. Once you’ve seen that its output is consistently good, switch it to autopilot. Going straight to autopilot on day one is how a small briefing mistake becomes a dozen published mistakes.

What runs needs abilities and an output

A schedule only matters if the agent can actually do the work and send it somewhere. Make sure it has the right abilities and a defined output destination before you set it loose.

Still need a hand?
Reach the team from inside the app, or get started free.
Start free trial