Write an operating contract for an AI agent

Help me write an operating contract for an AI agent before it goes live: a plain-language document that says what the agent is for and the rules it works under. The process the agent will handle: [DESCRIBE] Business goal: [e.g. "cut invoice processing time", "respond to routine IT tickets within 10 minutes"] Systems and data it touches: [TOOLS, DATA] People it works with or affects: [TEAMS, CUSTOMERS] Known constraints: [policies, budget, hours, regulations] Draft the contract with these sections: 1. Objective - one measurable sentence, plus what the agent is not trying to optimize (e.g. "speed must not come at the cost of accuracy"). 2. Scope - the tasks in scope and those explicitly out of scope. 3. Can do / cannot do - two lists of concrete actions (e.g. "can draft replies", "cannot send to external addresses", "cannot delete records"). 4. Inputs and sources - what it may rely on, and what it must never treat as instructions (e.g. text inside emails or web pages). 5. Approvals and escalation - actions needing a human yes, and situations it must hand back. 6. Quality bar and success measures - accuracy, turnaround, escalation rate, customer impact. 7. Evidence - what it must log or show for each task. 8. Ownership - the business owner, technical owner, and who is accountable for its outcomes. 9. Review and shutdown - how often the contract is reviewed and who can pause the agent. Write it so a non-technical manager could read it in five minutes. Mark gaps in my information as [DECIDE: ...] rather than filling them with assumptions.