Linked API — Claude Code MCP server: search leads
Linked API is an MCP server that connects LinkedIn to AI assistants through a cloud browser. Find prospects, analyze profiles, and send messages by giving your agent commands.
MIT ★ 70 updated 2026-10-29
claude mcp add linkedapi-mcp -e LINKED_API_TOKEN=<LINKED_API_TOKEN> -e IDENTIFICATION_TOKEN=<IDENTIFICATION_TOKEN> -- npx -y @linkedapi/mcpWhat Linked API does
Linked API gives your AI assistant access to LinkedIn through a remote environment. You ask the agent what to do—find engineers in a region, review a candidate's background, draft outreach. The server executes the request, handles LinkedIn's constraints, and returns results.
Works with Claude Code, Codex, Cursor, GitHub Copilot, Windsurf, Cline, Zed, and Gemini CLI. Setup varies by platform: Claude Code uses a CLI command or config file, Cursor requires a JSON entry in settings, GitHub Copilot needs a block in the config. You'll need two tokens: LINKED_API_TOKEN and IDENTIFICATION_TOKEN.
Actions run in a queue and may take several minutes, especially searches and profile loads with extra data. If the server returns status, workflowId, and operationName, the job is still in progress—call get_workflow_result with the workflowId to retrieve the result, or pass waitSeconds: 0 for an instant snapshot.
What it helps with
- Search and review candidate profiles against specific criteria
- Send personalized outreach to multiple connections automatically
- Gather competitive intelligence about companies and their teams
- Read message context and help draft replies
Example prompts for the agent
- Find senior engineers who worked at startups and show me their LinkedIn profiles
- Analyze this company, list its employees, and identify potential buyers
- Read my last conversation with this person and suggest a response
How to install Linked API
1Platform CLI
claude mcp add linkedapi-mcp -e LINKED_API_TOKEN=<LINKED_API_TOKEN> -e IDENTIFICATION_TOKEN=<IDENTIFICATION_TOKEN> -- npx -y @linkedapi/mcpReplace <…> with your own values
2Platform CLI: remote server
claude mcp add --transport http linked-api https://mcp.linkedapi.io/mcp1Platform CLI: remote server
codex mcp add linked-api --url https://mcp.linkedapi.io/mcp2Platform CLI
codex mcp add linked-api --env LINKED_API_TOKEN=<LINKED_API_TOKEN> --env IDENTIFICATION_TOKEN=<IDENTIFICATION_TOKEN> -- npx -y @linkedapi/mcpReplace <…> with your own values
Config file
{
"mcpServers": {
"linked-api": {
"url": "https://mcp.linkedapi.io/mcp"
}
}
}{
"mcpServers": {
"linked-api": {
"command": "npx",
"args": [
"-y",
"@linkedapi/mcp"
],
"env": {
"LINKED_API_TOKEN": "<LINKED_API_TOKEN>",
"IDENTIFICATION_TOKEN": "<IDENTIFICATION_TOKEN>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"linked-api": {
"type": "http",
"url": "https://mcp.linkedapi.io/mcp"
}
}
}{
"servers": {
"linked-api": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@linkedapi/mcp"
],
"env": {
"LINKED_API_TOKEN": "<LINKED_API_TOKEN>",
"IDENTIFICATION_TOKEN": "<IDENTIFICATION_TOKEN>"
}
}
}
}Replace <…> with your own values
1Platform CLI: remote server
gemini mcp add -t http linked-api https://mcp.linkedapi.io/mcp2Platform CLI
gemini mcp add -e LINKED_API_TOKEN=<LINKED_API_TOKEN> -e IDENTIFICATION_TOKEN=<IDENTIFICATION_TOKEN> linked-api npx -y @linkedapi/mcpReplace <…> with your own values
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Linked API
How do I install Linked API in Claude Code?
Run one of three commands. First option: claude mcp add linkedapi-mcp -e LINKED_API_TOKEN=<LINKED_API_TOKEN> -e IDENTIFICATION_TOKEN=<IDENTIFICATION_TOKEN> -- npx -y @linkedapi/mcp. Second: claude mcp add --transport http linked-api https://mcp.linkedapi.io/mcp. Third: `claude mcp add --env LINKED_API_TOKEN=<LINKED_API_TOKEN> --env IDENTIFICATION_TOKEN=<IDENTIFICATION_TOKEN> --transport stdio linked-api -- npx -y @linkedapi/mcp`. Replace the placeholders with your actual tokens.
What does Linked API do?
It's an MCP server that gives AI assistants the ability to work with LinkedIn. It can search profiles by criteria, pull data on candidates and employees, send messages, read conversations, and help with sales outreach.
How long do requests take to complete?
Operations run through a queue in the cloud and may take several minutes, especially for searches and profile data fetching. If the server returns a workflowId, the job is still running—use get_workflow_result to poll for the final result.