Skill
Build a targeted B2B prospect list with the leads-finder source.
Playbook for turning a vague 'find me customers' into a filter set the leads source can actually run: which seniority, department, industry, size, revenue and funding values exist, how to phrase job titles, and how to read filterNote instead of over-promising the list.
The full instructions this skill installs.
Use this when someone wants a list of real people to contact — prospects, buyers, decision-makers, a B2B list — rather than advice about prospecting.
The source is leads-finder. It searches a contact database, not the live web, and returns people with verified work emails.
Filter on the company first, the person second. A search for "CEO" returns every CEO of every corner shop on the continent. A search for CEOs at 11–50 person software companies in France returns a list somebody can actually sell to. Every filter you add is worth more than every extra row you fetch.
query and location are both mandatory. A job title with no place returns whoever the index feels like and the list is worthless. If the user has not said where, ask — do not guess a country.
| They said | Use |
|---|---|
| "decision makers", "the boss" | seniority: ["founder", "owner", "c_suite"] |
| "someone senior in marketing" | seniority: ["vp", "head", "director"] + team: ["marketing"] |
| "startups" | size: ["1-10", "11-20", "21-50"] + often funding |
| "funded startups" | funding: ["seed", "series_a"] |
| "enterprises", "big companies" | companySize: ["1001-2000", "2001-5000", "5001-10000"] + minRevenue |
| "agencies" | industry: ["marketing & advertising"] |
| "SaaS", "software companies" | industry: ["computer software"] |
query is the job title, and a comma-separated list means "any of these": "cto, vp of engineering, head of engineering". Titles vary far more than people expect — casting two or three is usually better than one.
seniority, team, companySize, funding, minRevenue and maxRevenue only accept the source's own values, and the tool schema lists them — stay inside it.
industry is a list of 148 fixed names. If what you want is not one of them ("fintech", "AI", "coworking"), pass it as companyKeywords instead. If you pass it as industry anyway the platform downgrades it to a keyword search and tells you so in filterNote — pass on that note to the user rather than letting them think an industry filter applied.
location accepts countries and "region, country" pairs ("california, us", "bavaria, germany"). A city that is not in the list is applied as a city filter, which is usually what was wanted.
emailStatus defaults to ["validated"]. Leave it there. Widening it to not_validated or unknown inflates the row count with addresses that bounce, and a bouncing list damages the sending domain. Only widen if the user asks for it explicitly, and say what the trade is.
Default 25. That is a real list somebody can work through in an afternoon. Go higher only when asked, and remember the run is billed per lead — 100 is the ceiling. Never fetch a large list "to be safe".
resultCount is how many rows came back.filterNote, when present, means a filter did not apply as asked. Always relay it.email, mobile, linkedin or companyName can genuinely be null. Say "not listed". Never fill one in.Name the best few by name, company and title even though a card is shown — never answer a search with only a count. Then hand over the spreadsheet (see the Lead List Export skill) and offer the obvious next step: narrow the filters, or draft the first outreach email.