6sense
Ask an AI agent about 6sense account insights, buying signals and keyword trends.
Search B2B contacts and companies, reveal verified emails and phone numbers, and pull buying signals.
Opens Agent Studio, where connecting is one click. The connector URL below works in any MCP client.
Paste it into any MCP client. Setup docs
The Lusha MCP server connects Lusha's B2B contact database to Claude, ChatGPT, Codex and any MCP-compatible agent. It exposes 24 tools that search for people and companies, reveal verified work emails and phone numbers, and return buying signals. The data is personal information about third parties, and revealing it deducts Lusha credits.
Verified connector
Listed by Anthropic as a partner connector in its Connectors Directory.
Connection checked by Agentman on .
Anthropic states this reflects the level of review a connector received, not a security audit.
Tool names from Anthropic's directory listing. This server requires sign-in, so we could not read tool descriptions or parameter schemas.
mcp scope, and the authorization server's extra api_access scope is a platform scope rather than a narrower alternative. Per-call approval in your client is the granular control.initialize on 2026-08-18, so names came from the directory snapshot and meanings from Lusha's documentation. No tool was called and no person was looked up.Yes, but only when you reveal contact data. Lusha charges 1 credit per email revealed and 5 credits per phone number, with search, filtering and company browsing at 0 credits. Its documentation states MCP calls consume API credits, and that API requests carry a 1-credit minimum even when no results come back.
A paid Lusha plan with available API credits. Lusha's setup article lists a paid plan and an Admin or Manager role as requirements, because generating an API key needs that role. Lusha sells Free, Starter, Pro, Premium and Scale tiers, and its Do Not Call filtering feature is restricted to Scale plans only.
Use the bare host with an API key header. Lusha's documentation gives OAuth for Claude, ChatGPT and Codex only, and tells every other client — Cursor, VS Code, n8n, Gemini CLI — to point at https://mcp.lusha.com and pass an x-api-key header. Lusha notes the header name must be lowercase or authentication fails.
Third parties who never gave Lusha their data directly. Lusha aggregates business contact details from public professional profiles, company websites, business directories and contributed sources. It states it relies on legitimate interest as its legal basis and does not collect opt-in consent from individuals for third-party outreach.
Through Lusha's removal form or its privacy team. Lusha documents an automated removal form and a privacy address, and states submitted details are permanently deleted. Opted-out contacts are suppressed from search, enrichment and any Lusha output. Bulk removal through the form is not available, so multiple records need the privacy team directly.
No — opt-out reporting runs outside the connector. Lusha sends CSV opt-out reports to a designated account Admin, typically monthly, covering only contacts who opted out since the previous report. Lusha also offers an opt-out webhook for real-time notification. Building a full suppression list means aggregating every report you have received.
It can spend credits, and Anthropic's snapshot names a purchase_options tool. Reveals deduct credits automatically, and Lusha documents an in-conversation top-up flow on its Claude cards. Lusha lets Admins cap credits per user or group on Premium and Scale plans, which is the control that limits an agent's spend.
Twenty-five requests per second on general endpoints. Lusha's API documentation adds 5 requests per minute for the Account Usage API and 100 per minute for the Webhooks API, and returns remaining quota in response headers such as x-rate-limit-minute. Exceeding a limit or a daily quota returns HTTP 429.
.md appended). docs.lusha.com/robots.txt is a blanket Allow: / with no Content-Signal line, and the host publishes both an llms.txt index and a sitemap.xml. · retrieved 2026-08-18llms.txt; found via the sitemap. · retrieved 2026-08-18https://mcp.lusha.com/mcp/claude — initialize returned/mcp form and the host root (all HTTP 200, 2026-08-18). Each names scopes_supported: ["mcp"], authorization_servers: ["https://auth.lusha.com"] and itself as the resource. · retrieved 2026-08-18scopes_supported: ["mcp", "api_access"], PKCE S256 and a registration_endpoint. The copy served from mcp.lusha.com lists only the mcp scope. · retrieved 2026-08-18permissions value and has_mcp_app (snapshot dated 2026-08-16) · retrieved 2026-08-16www.lusha.com. Its robots.txt carries Content-Signal: search=yes, ai-train=no, use=reference and disallows ClaudeBot by name, and the host returns HTTP 403 to a command-line client. Lusha's privacy notice, data page, trust centre and removal form live there; this page links to them and synthesises nothing from them. Everything above comes from docs.lusha.com, which sets no such signal.Connect Lusha once and your agents call these tools on their own: on a schedule, in a workflow, with nobody at the keyboard.