Run an Agent: manual, schedule, or webhook
In 配備と実行 (Deploy & Run), place an approved Agent in a Room and choose when it can run. Start manually, then add a schedule or webhook only after you have checked real results.
| Run method | Best for | First recommendation |
|---|---|---|
| Manual | Meeting follow-ups or important deal checks | Start here |
| Schedule | A daily stalled-deal review or weekly executive summary | Use after manual runs are stable |
| Webhook | Work initiated by an event in an external system | Use after signature testing and permission review |
Manual run
- In 配備と実行 (Deploy & Run), choose the approved Agent and its Room.
- Check the input and reference scope.
- Select 手動実行 (Manual run).
- Read the output, references, and failures in 実行レビュー (Run Review).
Schedule
Use a schedule for recurring work. An internal Room editor chooses the Room, time, time zone, and definition. Creating a schedule does not make it active by itself; activation is a separate action that begins future automatic runs.
A schedule does not run an Agent across every Room. It runs only within the selected Room and permitted scope.
Webhook
A webhook starts a run from an external event. Only a verified signed event is accepted. Webhook creation and activation are Workspace-admin actions, so ask an administrator before you plan a webhook-based workflow.
Store the webhook secret only in an approved secret store when it is created. Do not paste it into a prompt, Room material, Run Review, or support request.
Related articles
- Evaluate, approve, and deployCheck expected results in simulations before deploying an Agent to a Room
- Rooms, sources, connections, and permissionsUnderstand what an Agent can reference and what to check before enabling MCP or write actions
- Troubleshooting Agent StudioCheck goals, evaluation, deployment, and run results when an Agent does not work as expected