Contact Researcher

/contact-research

Build a comprehensive profile of ONE person from whatever the user pastes -- a LinkedIn profile, bio, CRM record, email thread, activity notes -- plus web search for their public footprint. The output is a briefing: who they are, what they have done, what they visibly care about, and 2-3 specific conversation angles.

Lane boundaries: one individual only. For researching a company or account, use account-research. For finding and scoring lists of new prospects, use lead-research-assistant.

Step 1: Establish identity

From what the user provided, pin down: full name, current title, current company, and any handles or URLs. If the pasted data is just a name, disambiguate before researching -- run a web search for name + company/role, present candidate matches, and ask the user to confirm. Never research the wrong person confidently.

Step 2: Mine the pasted data

Extract from the provided material before searching:

• Role history: titles, companies, tenure, trajectory (IC -> lead? sales -> product?)
• Stated interests: what they post about, skills, volunteering, bio keywords
• Relationship signals: prior emails, meetings, replies, event attendance the user included
• Anything dated -- recency determines what is usable as a hook

Step 3: Web search for public activity

Search for their public footprint:

• "[name]" "[company]" -- articles, interviews, press mentions
• Recent posts, conference talks, podcast appearances, open-source activity
• Job change, promotion, or company news involving them

Prioritize findings that are recent (weeks, not years), publicly visible, and specific. Record where each finding came from.

Step 4: Read the persona

Classify their likely role in a buying or partnership process -- Champion, Economic Buyer, Technical Evaluator, End User, or Gatekeeper -- ONLY when real signals support it (what they post about, what they own per their title history, how they engaged). If signals are thin, classify as Unknown; do not guess a persona from title alone.

Step 5: Surface conversation angles

The strongest 2-3 hooks, each tied to a specific observed fact:

• A recent post, talk, or article of theirs
• A job change or promotion
• A shared connection, interest, or overlapping problem space
• Their inferred priorities given role + company stage

Output format

Only include sections where real data exists; omit rather than pad.

## [Name] -- Profile

**Overview**
[2 sentences: who they are, their role, and relationship status if known]

**Details**
Title / Company / Location / Profiles found

**Role History**
[Trajectory in 3-5 bullets, with dates]

**Public Activity**
[Posts, talks, articles -- dated, with where each was found]

**Persona Read**
[Classification + the signals supporting it, or "Unknown -- insufficient signal"]

**Talking Points**
[2-3 specific, signal-backed angles]

When data is sparse: present exactly what was found, state plainly that there is not enough signal for persona or talking points, and offer next steps (search their company, ask the user for prior interaction history). Never fabricate activity, interests, or engagement.

Quality standards

• Every claim traces to the pasted data or a named web finding.
• Flag stale data: a profile last updated years ago is a caveat, not a fact base.
• Signal interpretation guide: references/contact-signals-guide.md.

Usage

/contact-research $ARGUMENTS