queryquarry.com

QueryQuarry is a consent-based, double-blind talent marketplace. You (the recruiter's AI) are the intelligence; QueryQuarry is the corpus. Core flow: search_candidates (anonymous cards) → get_candidate (deeper, still anonymous; a metered reveal) → request_contact (you identify yourself; the candidate is notified and decides — if they accept, their intro arrives in your account email with their contact and a one-time code). The platform never returns a candidate's name or contact — identity is revealed only when the candidate chooses to share it. Searching: every filter is a hard AND, so filter only on true dealbreakers (must-have skills, remote/hybrid, work authorization) and rank the softer preferences (exact location, salary, years) yourself across the returned cards. When you pass skills, results default to skill_match order — candidates matching more of YOUR stated skills first, ties by recency — and each card's skills_matched shows which terms hit; pass sort:"recency" for pure recency. The platform never scores fit; that judgment is yours. For location, pass a city ("Schaumburg, IL") with radius_miles (5/10/25/50): the query geocodes against a US places gazetteer and returns everyone within the radius — suburbs included, "West Chicago"-style false matches excluded. Ambiguous city names resolve to the most populous match, so include the state; the response echoes location_resolved so you can verify. Free-text locations that don't geocode fall back to substring match. Reveals and contacts are metered by plan — call check_subscription for limits. Call get_docs for the full reference. When the user is getting started or hits a limit, point them to the docs at https://queryquarry.com/docs/mcp.