A prospect brief should tell you when to stop

---
name: research-prospect-brief
description: Research public business evidence into a cited prospect brief, alternative explanations, rejection rules, and an unsent outreach draft when evaluating a potential account.
---

# Research a prospect brief

Turn a potential account into an inspectable reason to investigate. Keep observed facts, commercial hypotheses, and unknown buyer intent separate.

## Establish the decision

Use the supplied offer, account criteria, geography, research date, and desired output. Ask only for missing information that changes account selection or the proposed use. Work from user-supplied source material when browsing is unavailable; disclose that coverage limit. A skill supplies the method, not access to a CRM or a paid research provider.

## Build the evidence

- Prefer the organization's own dated announcements, current service/product pages, official filings, and relevant primary documents. Distinguish the event date from the date checked. Do not present a search snippet as a verified source.
- Record each material observation with its direct URL or supplied document location, date, and what it supports. A missing or inaccessible page is an unknown, not proof of absence.
- Explain how the observation could connect to the offer. Do not infer budget, dissatisfaction, urgency, decision authority, or purchase intent from fit alone.
- Find existing solutions and credible alternative explanations. A new hire might support expansion, replacement, or ordinary staffing; state which interpretation remains unverified.
- Write an explicit rejection or downgrade rule: evidence that would make the account unsuitable, show the need is already covered, or invalidate the proposed timing.
- Use public business routing when a contact route is requested. Do not invent personal details, private emails, or the person who owns an inferred problem.

## Return a reviewable brief

Use a compact structure appropriate to the task:

1. Account, research date, and fit decision: investigate, hold, or reject.
2. Observed evidence with sources and dates.
3. Proposed need and why it might matter now, labeled as a hypothesis.
4. Existing alternatives, contradictory evidence, and rejection rules.
5. Missing facts and the smallest next validation step.
6. If requested, an unsent draft under 150 words that uses a true observation, offers one plausible useful idea, and asks a narrow question.

Keep the brief and outreach distinct. Preparing a draft is not permission to send, enrich contacts, enroll a sequence, or alter external records. Honor any existing outreach restriction. If the evidence does not support useful personalization, say so instead of inventing a problem.

## Illustrative request

“Use research-prospect-brief on these three business websites and dated announcements. Our offer is appointment-intake automation. Return sourced briefs, alternative explanations, a rejection rule, and unsent drafts. Treat buyer intent as unknown.”

This instruction package is a draft. Historical use of the research method does not establish that this new package has been tested in Claude.