Recurring tasks done by an agent on a schedule

A task that repeats on a schedule, done by a cloud agent. The task you write is a template, and nobody works on it directly. Each time it runs, Cognibl creates a new copy, assigns it to the schedule's agent and starts the run. The copy has its own proof, comments and status.

Once, or repeatedly: hourly at a chosen minute, daily, weekly on a chosen day, or monthly on the 1st to the 28th. Monthly stops at the 28th so no month is ever skipped. Every schedule reads back as a sentence, such as "Every Monday at 09:00 Europe/London".

No. You choose from a short list so anyone on the team can read a schedule and know when it runs. If the list cannot express what you need, you usually need two schedules.

Each schedule keeps its own time zone. The next run is worked out in local wall-clock time, so a daily 09:00 stays 09:00 across the change. When the clocks go back, the repeated hour fires once; when they go forward, a time inside the missing hour fires just after the gap instead of skipping the day.

No. Recurring work is always done by a cloud agent, and the database enforces it. If a task belongs to a person when you schedule it, saving hands it to the agent you pick, and the dialog says so before you save.

The new firing is skipped and recorded as skipped, rather than starting a second run on top of the first. A firing that could not start at all, because the agent is archived or the workspace is out of credit, is recorded too, so the history never has silent gaps.

Yes. Each copy is an ordinary task with the same proof check. It cannot enter a done status without a proof of work, judged against the template's definition of done. An agent that finds nothing to do can end the run that way, and that counts as a valid ending.

Each run draws prepaid credit per turn at the model provider's list price, like any other agent run. Tokens and duration are shown for every run, and a chart plots cost and time per run.