LeadShark MCP

Overview

The LeadShark MCP (Model Context Protocol) server lets AI assistants like Cursor, Claude, and ChatGPT manage your entire automation workflow through conversation.

What You Can Say

Talk to your AI assistant naturally. It figures out which tool to use.

Tools (40)

Complete automation control. Each tool does one thing well.

marks an Apex-only tool. Everything else is available on Pro+ and Apex.

The core “Act” arm of the agent loadout — engage as yourself with the posts and people you surface anywhere across the toolkit: discover_lead_magnets and list_signals, your own feed, get_person_activity (the posts someone's recently been commenting on and reacting to), list_post_engagers (everyone who engaged a given post), list_person_posts / list_recent_posts, or search. Use comment_on_post / react_to_post with a post's post_id, and send_connection_request to invite a person to connect. Go further into conversations with send_message (DMs, or InMail a non-connection with as_inmail) and engage_with_comment (reply to or like a comment), and read what's happening with list_recent_messages, get_messages_with_person, feed (your own home feed), and get_lead_activity (a lead's full history with you). Acting as yourself is Apex; the Company-Page voice equivalents below are Apex too.

Act as a LinkedIn Company Page you administer. Start with list_companies to get the page's id, then pass it as organization_id to the other tools. Pro+ covers reading the pages you manage (list_companies) and scheduling / reading a page's posts (schedule_post_with_automation, list_recent_posts). Apex covers acting in the page's voice — comment_as_company, react_as_company, engage_with_comment (as company), and manage_company_reply_automations (auto-reply to comments on a company post). You must be an admin of the page on your connected LinkedIn account.

The SEE and DECIDE half of the agent loop. search finds ICP-fit people or companies on LinkedIn by keyword and degree (proactive prospecting); list_person_posts pulls a person's recent posts so you have something concrete to act on; get_person_activity surfaces recent comments and reactions; list_post_engagers turns a post into a warm lead list; discover_lead_magnets pulls fresh prospect-rich posts; and enrich / enrich_company supply decision-ready profile data. All require Apex except Discover and base enrichment, which are available on Pro+ and Apex. enrich's include_email stays Apex and uses the linked PersonEnrich wallet.

Work the leads your lead magnets captured. list_lead_pages lists your Pages (lead-magnet / quiz pages) with per-page capture stats; get_page_leads pulls the actual captured emails for one page (deduped and enriched with the LinkedIn profile when known); manage_lead_magnet_links lists, creates, and edits the leads.sh capture links you have in play — so the full post → capture link → automation flow works without opening the app. This is the bridge from a lead magnet into outreach — feed a captured lead's linkedin_url into enrich then send_connection_request / send_message. (This surface feeds the Apex outreach loop, so it is Apex.)

Image Attachments

Attach images to your scheduled posts by providing a publicly accessible URL. LeadShark downloads and stores the image securely.

Example Workflow

A typical conversation from discovery to optimized automation:

Setup

Requirements

• • Node.js 18+
• • LeadShark Pro+ or Apex subscription
• • API key from Dashboard Settings

Install

npm install @leadshark/mcp-server

Cursor Configuration

Create .cursor/mcp.json:

{
  "mcpServers": {
    "leadshark": {
      "command": "npx",
      "args": ["-y", "@leadshark/mcp-server"],
      "env": {
        "LEADSHARK_API_KEY": "${LEADSHARK_API_KEY}"
      }
    }
  }
}

Restart Cursor after adding.

Claude Desktop

claude mcp add leadshark \
  --transport stdio \
  --env LEADSHARK_API_KEY=your_key \
  -- npx -y @leadshark/mcp-server

curl -sL https://apex.leadshark.io/skill/openclaw | sh

Safety Model

Settings Reference

All available settings you can configure via the MCP:

Connect to Claude.ai

The fastest way to use LeadShark with Claude. No install, no API key — Claude.ai registers itself with the LeadShark MCP server automatically over OAuth.

https://apex.leadshark.io/mcp

1. 1

   Open Customize → Connectors

   In Claude.ai, open Customize from the left menu, then Connectors (it recently moved out of Settings). Click the + next to “Connectors” and choose Add custom connector.

   

   

2. 2

   Name it & paste the MCP URL

   In the dialog that opens, give it a name like LeadShark Apex, then paste https://apex.leadshark.io/mcp as the URL. Leave the optional OAuth fields blank — Claude registers itself automatically — and click Add.

   

3. 3

   Click Connect

   LeadShark Apex now shows in your Connectors as “You are not connected to LeadShark Apex yet.” Click Connect to start the sign-in.

   

4. 4

   Sign in & authorize on the LeadShark page

   Claude opens a popup to apex.leadshark.io. Sign in with your LeadShark account, review the requested scopes (mcp:read and mcp:write), and click Approve.

   

5. 5

   You're connected

   Claude lands back on the Connectors page with LeadShark · Configure visible in the list.

   

6. 6

   Set tool permissions

   Open LeadShark · Configure. Claude lists every Apex tool and lets you choose when it can run them. Set the read-only tools (the ~23 that only look — discover, enrich, list, read) to Always allow so reads run instantly. Leave the write & delete tools (the 15 that act on your account — comment, connect, message, etc.) on Ask every time (the hand “needs approval” icon) so Claude checks with you before each action. Then try a prompt like “Show me my last 3 LinkedIn posts” to confirm it works.

   

Connect to ChatGPT

Add LeadShark to ChatGPT as a custom app. No install, no API key — ChatGPT registers itself with the LeadShark MCP server automatically over OAuth.

https://apex.leadshark.io/mcp

1. 1

   Enable Developer mode & create an app

   In ChatGPT, open Settings → Apps and turn on Developer mode (under Advanced) — this is required to add a custom connector. Then click Create app to open the New App dialog.

   

   

2. 2

   Add the MCP server

   Give it a name (e.g. LeadShark Apex), set Connection to Server URL, and paste https://apex.leadshark.io/mcp. Set Authentication to OAuth, tick I understand and want to continue, then click Create.

   

3. 3

   Sign in with LeadShark

   ChatGPT opens an Add LeadShark Apex to ChatGPT screen. Click Sign in with LeadShark Apex to start the OAuth flow.

   

4. 4

   Authorize & you’re connected

   On the apex.leadshark.io consent screen, review the requested scopes (mcp:read and mcp:write) and click Approve. Try a prompt like “Show me my last 3 LinkedIn posts” to confirm it works.

   

Connector icon

When you add LeadShark Apex as a connector in Claude or ChatGPT, it asks for an icon. Click any to copy it straight to your clipboard, or download the PNG — all 256×256, a few KB each.

Copy

LeadShark · Primary

Copy

Chess queen · Primary

Copy

LeadShark · Black

Copy

Chess queen · Black