Brief a research agent to prep you for a sales call

About this exercise

You have a call tomorrow and want an AI research agent to pull talking points from the sources below. Write its brief: use only sources from 2025 or later, never present a rumor as fact, give exactly 4 notes of under 20 words each, each tagged with its source like [S1], and never email or connect with anyone at the company. Output exactly 5 numbered plain-text lines: 4 notes, then DONE.

Common mistakes when you ask AI for this

• Not showing the log format. Left alone, the AI writes a friendly report with headings instead of a checkable action log. Try: Show it literally: "Format every line like 1. ACTION: tool(args), then ASK_HUMAN: and DONE: lines. Plain text."
• No length or sourcing rule. The AI writes long notes with no way to check where a claim came from. Try: Say: "4 notes under 15 words, each ending with its source tag like [S1]."
• Trusting stale or shaky sources. The AI repeats a 2022 contract and an anonymous rumor as current facts. Try: Add: "Only sources from 2025 on. Never state a rumor as fact."
• Letting it reach out. Given email and LinkedIn actions, the AI warms up the contact for you. Try: Add: "Never email or connect with anyone."

What a strong prompt includes

• Sets a date cut-off
• Handles rumors
• Asks for source tags
• Forbids outreach
• Shows the action-log format

Practice this now ↑

Related exercises

• Alert sales to a competitor's price change, facts only (Research & AI Search · Product managers)
• Approve or return expense claims against the policy (Productivity AI · Finance and accounting)
• Approve part of an overtime request with the math (Writing AI · Managers and team leads)
• Ask the right questions about a usage drop (Customer Support & Success AI · Customer success managers)
• Assign accounts to segments by the first matching rule (Sales AI · Revenue operations)