Skip to main content
Building and saving agents is broadly available. Running one is the paid part, and running one on its own is what this page covers. Two tabs handle it: the Automation tab, where you set an agent to run on a schedule, from a webhook, or across a batch of rows, and the Scheduled tab, where every run those settings produce shows up on a dated timeline.
Agents are Beta and review-gated. Automation does not change that. Nothing is published to your site, emailed, or sent to any outside service without a person approving it first, and any step that would reach out stops for approval before it runs. Treat what an agent produces as a draft to review, not an answer to trust unattended.

​
What can run on its own

Only a published custom agent can be automated. A recipe or a draft can’t. Publish the agent first, then it appears in the Automation picker. (Agents are also called “workflows” in a few places in the app; they’re the same thing.) Automation always uses the fixed published version. If you edit the draft later, a live schedule or webhook keeps running the version you published until you publish again. That’s deliberate, so a half-finished edit can’t quietly change a job that’s already running.

​
The Automation tab

Pick a published agent, then set up one of three ways to run it. Each run keeps its own record, whichever way it started.
1

Schedule

Run once or on a repeating cadence: Doesn’t repeat, Hourly, Daily, Weekdays, Weekend, Weekly, Monthly, or Annually, at a time in your local timezone. You can pass optional inputs for the run. After it’s saved, each schedule shows its next run and its latest result, and you can edit it, pause and resume it, or delete it. Pausing stops future runs; runs already started aren’t cancelled and past results stay available.
2

Webhook trigger

Give another product a secret URL it can call to start the agent. Name the trigger, create it, and copy the URL then — it’s shown once, at creation, and never again. Repeat calls can carry an idempotency key, so if the same call arrives twice the agent doesn’t start the run twice. Each trigger lists when it was last used and links to its last run, and you can revoke it whenever you like. Revoking takes effect immediately; saved runs remain.
3

Bulk run

Run the agent once per row across many rows at once. Give the batch a name, then supply the rows one of three ways: upload a CSV (up to 8 MB and 10,000 rows), connect a Google Sheet by link, or paste a JSON array. The first row of a CSV or sheet must match the agent’s input names. A connected sheet has to allow anyone with the link to view it, and FixAEO never edits the sheet — it only reads it. The batch then shows row-by-row progress with a status per row and a link into each run.
A webhook URL is shown once, at the moment you create it. Copy it then and store it somewhere safe. If you lose it, there’s no way to see it again — you revoke the old trigger and create a new one.

​
The Scheduled tab

The Scheduled tab is the dated timeline of runs your automations produce. It shows the runs coming up over roughly the next 30 days and the last 90 days of past ones, grouped into Past, Today, Tomorrow, This Week, Upcoming 30 Days, and Later. Each run carries a status:
  • Scheduled — set to run and waiting for its time.
  • Running — in progress now.
  • Awaiting review — paused at a review step for a person to approve or reject.
  • Succeeded — finished.
  • Failed — the run didn’t complete.
  • Missed — its time passed without the service starting it (for example, it was offline). Missed runs aren’t backfilled.
  • Cancelled — you cancelled it before it ran.
From a run you can open it to see its output and evidence. You can cancel a single upcoming run without touching the rest of the schedule, or remove the whole schedule so no future run happens. Removing a schedule leaves any run already started alone, and every past result stays available.

​
The weekly brand-health report

The Scheduled tab also manages a simpler, built-in schedule: a weekly brand-health report that compares this week to last and saves the result in the app. It’s the one automation you don’t have to build — start it with New weekly report. A few things set it apart from a custom agent’s schedule:
  • It’s weekly only, on a day and time you pick.
  • Its time is in UTC, so the schedule stays put across daylight-saving changes. (Custom-agent schedules use your local timezone instead.)
  • Every result is saved in the app. Nothing is emailed or published.
  • You can save up to 20 weekly report schedules per account. Paused ones still count toward that.
You can pause, resume, edit, or remove a weekly report the same way as any schedule.

​
What you need to run

Enabling or running any automation needs an active paid plan. Without one, FixAEO tells you plainly: “An active paid plan is required to start a workflow. Your saved results remain available.” You can still pause a schedule without a paid plan — pausing never starts a run. A few limits apply to every account, and they cover manual, scheduled, and automated runs together:
  • Up to 20 runs per rolling day.
  • Up to 3 runs active at once.
Automated runs draw on your AI Marketer work credits, the same pool the chat and manual runs use — see Plans and limits and Usage for the allowance and your remaining balance. If an agent is shared with you, changing its automation needs editor access; a viewer can follow the runs but not change the schedule.

Agents

Build, test, and publish an agent — the step before you can automate it.

Resources and review

The approval flow that every run pauses at before anything goes out.

Templates

Ready-made agents to start from, including the weekly brand-health report.

Plans and limits

Which plans can run agents, and the run and schedule caps.

Usage

The credits automated runs draw on, and your remaining balance.

AI Marketer

The conversational sibling that drafts work for you to review.