Mission Control is the task automation center in Yao Agents. You build Robots — autonomous workers that run on a schedule or respond to triggers — and they get things done without you asking every time. > You're the commander. Define the goal, assign the team, let it run. ## Why Mission Control exists AI Experts are powerful, but they're reactive — they work when you ask. If you need the same thing done every morning, every hour, or whenever something happens, you'd have to ask every time. That doesn't scale. Mission Control flips the model. You describe a job once: what the Robot does, what resources it has, and when it should run. After that, it handles itself — collecting, processing, writing, sending — and reports back with the results. ## What a Robot can do - **Work on a schedule** — daily briefings, weekly reports, hourly data checks. Set it once. - **Respond to events** — run when triggered by a Webhook, an incoming message, or an API call. - **Accept manual tasks** — click **Assign Task**, describe what you need, and Robot Host confirms the goal before the pipeline starts. - **Deliver to where you work** — send results by email, post to Telegram, push to Discord, or call an API. The output goes where you need it. - **Coordinate a team of AI Task Agents** — each Robot runs a Pipeline: specialist agents handle planning, research, execution, validation, and delivery in sequence.  ## How it works When a Robot runs — on schedule or manually triggered — it executes a **Pipeline**: a fixed sequence of stages, each owned by a specialist AI Task Agent. ``` Inspiration → Goals → Tasks → Validation → Delivery ``` Each stage passes its output to the next. The Robot Host stage handles the human interaction side — when you manually trigger a Robot, it first clarifies the task with you before the pipeline kicks off. The default agents for each stage work well without configuration. You can swap any of them for a custom agent if you need tighter control over a specific step. → [Replacing Stage Agents](/docs/en-us/mission-control/execution-pipeline/replacing-stage-agents) ## The difference between a Robot and an AI Expert | | AI Expert | Robot (Task Agent) | |---|---|---| | **How it starts** | You open a chat and ask | Schedule, event, Webhook, or manual trigger | | **Who drives it** | You, in real time | Autonomous — runs without you | | **Best for** | One-off tasks, exploration, questions | Recurring work, automated workflows | | **Output** | Conversational response in chat | Structured result delivered to a destination | For everyday tasks and questions, use AI Experts. For anything you need done regularly or automatically, build a Robot. ## What's next - [Creating a Robot](/docs/en-us/mission-control/creating-a-robot) — set up your first Robot in a few minutes - [Execution Pipeline](/docs/en-us/mission-control/execution-pipeline/pipeline-overview) — understand the stages a Robot runs through - [Scheduling & Triggers](/docs/en-us/mission-control/scheduling-and-triggers) — configure when and how a Robot runs - [Progress & Results](/docs/en-us/mission-control/progress-and-results) — read logs, check status, and access deliverables
What Is Mission Control
feedback
反馈