AI Agent Prompt Examples for Teams

Marketing agent prompt

Use this structure when the task needs audience thinking, messaging options, and a campaign-ready output.

Act as a launch strategist. Turn the product notes into three campaign angles for [audience]. Include target pain point, promise, proof needed, channel fit, and risks.
Act as a content planner. Build a two-week content plan from this offer. Include audience intent, post goal, angle, CTA, and review notes.
Act as a positioning analyst. Compare these customer pain points and recommend a messaging hierarchy with assumptions labeled clearly.

Use this structure when the agent needs to organize evidence, synthesize findings, and call out uncertainty.

Act as a research analyst. Synthesize these notes into findings, caveats, open questions, and recommended next steps for [decision].
Act as a market researcher. Group these competitor observations into themes and identify what evidence is strong, weak, or missing.
Act as an evidence reviewer. Check this draft for unsupported claims, missing context, and places where the conclusion is too broad.

Use this structure when the task needs decisions, tradeoffs, and implementation-ready output for a product team.

Act as a product operator. Turn this feature idea into a brief with user problem, scope, risks, launch tasks, and acceptance checks.
Act as a feedback analyst. Cluster these customer notes by theme, impact, frequency, and suggested next action.
Act as a release planner. Create a launch checklist for this feature with owner roles, dependencies, risks, and review gates.

Yes. Treat them as starting points, then replace the audience, context, output format, and checks with your task details.

Roles help the agent use the right perspective and produce output that fits the task.

For multi-step work, yes. Acceptance checks make the result easier to review and improve.