LinkedIn connections for AI agents

Give your agent your shared network.

Import LinkedIn connections from you and your friends. Your agent researches target companies, then uses 2ndhop to find warm paths.

Stays on your Mac Works with Claude & Codex

Why 2ndhop

Agents should not browse LinkedIn one profile at a time.

Today they control a browser, open pages, and read screenshots. It is slow, brittle, and hard to scale.

Browser controlScreenshotsOne profile at a time

2ndhop gives agents structured access to the connections you already exported.

Your agent finds the right companies. 2ndhop shows who you or your friends know there. Three steps.

1. You and your friends export your LinkedIn connections, then you import the CSVs into 2ndhop

2ndhopPeopleAccountsIntegrations7,262 connections

Import connections

Owner name

Priya Nair

✓Priya-Connections.csv4,182 connections ready

Import

Owners 3

2. Connect your agent to the shared network

2ndhopPeopleAccountsIntegrationsMCP · on

Connect an AI agent (MCP)

http://127.0.0.1:4762/mcp · loopback only

claude mcp add --transport http 2ndhop http://127.0.0.1:4762/mcp

Claude or Codex can now ask 2ndhop who you or your friends know at the companies it researched, then map the path to an introduction.

3. Ask Claude to research your ICP and find warm paths through 2ndhop

Claude

Find warm paths

Research fintech companies that match our ICP. Then use 2ndhop to find people that I or my friends know at the strongest matches.

I found 12 companies that match your ICP. I'll check the strongest candidates against your shared network.

Acme CorpGlobexInitechUmbrella

2ndhopmatch_targetsCompleted

companies["Acme Corp", "Globex", "Initech", "Umbrella"]

3 warm accounts foundAcme Corp via Priya and Sam · Globex via Maya · Initech via Jordan

Best path: Priya or Sam can introduce you to Dana Kim, VP Sales at Acme Corp.

Reply to Claude...2ndhop connected