pip install mcp

python server.py

{
  "mcpServers": {
    "linkedin-outreach": {
      "command": "python",
      "args": ["/path/to/linkedin-outreach-mcp/server.py"]
    }
  }
}

Tool: generate_connection_request
Input: {"name": "Sarah Chen", "title": "VP of Engineering", "company": "Stripe", "shared_interests": "AI infrastructure, developer tools"}
Output: 3 personalized variants, all under 300 characters

Tool: generate_inmessage
Input: {"recipient_name": "James Park", "recipient_title": "CTO", "goal": "book a demo of our observability platform", "tone": "sales"}
Output: Value-first sales message with soft CTA

Tool: generate_post
Input: {"topic": "AI in sales enablement", "key_points": "personalization at scale, shorter sales cycles, better lead scoring", "style": "thought-leadership"}
Output: Engaging post with hook, numbered points, CTA, and hashtags

Tool: analyze_profile
Input: {"profile_text": "VP of Engineering at Stripe. Building payments infrastructure for the internet economy. Previously Google, MIT CS. Passionate about developer experience and distributed systems."}
Output: Industry (Technology), Seniority (VP/Executive), pain points, conversation starters, interest areas

Tool: generate_outreach_sequence
Input: {"target_name": "Maria Rodriguez", "target_title": "Head of Talent", "target_company": "Figma", "goal": "partner on recruiting automation"}
Output: 5 messages with timing (Day 1 -> Day 25), each with strategic context

Tool: generate_comment
Input: {"post_content": "Unpopular opinion: cold outreach is dead. The future is warm introductions and community-led growth. Here's why..."}
Output: 2-3 insightful comments that add a new perspective, not generic agreement