Scheduling

Scheduling runs an agent without a human kicking it off — overnight reports, periodic scans, follow-ups that should fire hours after a customer signs up.

Two kinds, set on the board item:

• ONCE — runs at next_run_at, then stops.
• INTERVAL — runs at next_run_at, then repeats every interval_seconds.

API flow

Each scheduled task is a board item under an actor project. If you don't already have a project for the agent, create one first — see the agent runtime reference.

wacht api call createProjectTaskBoardItem \
  --param project_id=<project_id> \
  --body @task.json

task.json:

{
  "title": "<short label>",
  "description": "<the prompt the agent runs on each fire>",
  "schedule_kind": "INTERVAL",
  "interval_seconds": 86400,
  "next_run_at": "<first fire time>"
}

description is the prompt the agent reads on every run. next_run_at is the first fire time; subsequent INTERVAL runs are queued at last_run + interval_seconds.

Run wacht api describe createProjectTaskBoardItem for the full payload — including the optional mounts array for attaching sandbox-storage files into the thread filesystem on each run.

Stopping a schedule

wacht api call archiveProjectTaskBoardItem \
  --param project_id=<project_id> --param item_id=<item_id>

In-flight runs complete; no further runs are queued.