Audits on a schedule
Run the auditors automatically and get the report — without opening Claude, and without remembering to.
Set it once
Choose the auditors and the cadence in Settings. Runs continue on their own and pick up where they left off on a big site, rather than dying at a gateway timeout and leaving you with half a report.
The report
Exportable, and led by movement — what changed since last time, section by section: failures, warnings, how many pages were checked, what was fixed and what is new. The site score and its arithmetic live on the dashboard.
Still your switches
The same safety switches that decide what Claude may touch apply to a scheduled run. A read-only site stays read-only whether the run was started by you or by the clock.
How it works
Choose the audit groups and the cadence; runs execute in background chunks on your site’s scheduler and save findings to the same tabs a manual run uses. Finished runs record the Autopilot Site Score, building the history that shows a site getting better — or catching the week it got worse.
Who this is for
Anyone with a maintenance contract to honor. The schedule keeps the baseline fresh, so the monthly white-label client report opens with a real diff instead of a scramble — and a regression from a plugin update surfaces on your dashboard, not in a client email.
Cadence beats heroics. A weekly audit that takes zero of your minutes catches the broken heading, the ballooning image and the missing meta description while each is still one finding, not a project — and long gaps between checks are exactly where small regressions compound into rescue jobs. Set the schedule once per site, then let the score history do the arguing at renewal time. Findings from scheduled runs land in the same tabs and the same client reports as manual ones, so the schedule changes when the work happens — never where it lives.