What is MCP Server for LinkedIn?
MCP Server for LinkedIn is an independent open-source project that exposes LinkedIn actions to MCP-compatible AI agents. It supports profile and company research, people and job searches, feed reading, connection requests, and conversation management. It interacts with LinkedIn through the user's browser session rather than requiring a LinkedIn developer application.
The server can run through uvx, Docker, or a source installation. It supports local stdio communication and Streamable HTTP for web-based MCP clients. Authentication can reuse a supported local browser's signed-in session or open a browser for manual login. Tool calls share a browser session and are serialized, so concurrent clients may need to wait or retry.
The project is not affiliated with LinkedIn or Microsoft. Its operation depends on authentication and LinkedIn's page structure. Message sending and replies require confirmation, while the Easy Apply tool only reads application forms and does not fill in or submit them.
Key capabilities
- Profile research Read profile sections and search for people using keywords, location, connection degree, and current company.
- Company research Retrieve company information, recent posts, open jobs, and employee search results.
- Job discovery Search jobs, inspect saved postings and job details, and read application destinations or Easy Apply forms without submitting them.
- Conversation management Read and search inbox threads, archive conversations, and send messages or replies after confirmation.
- Network and feed access Send or accept connection requests, read the home feed, and search posts.
- Session management Reuse a supported browser login, check session status, and close the shared browser without deleting the saved login.
Where MCP Server for LinkedIn fits
The server fits agent-assisted candidate research, company investigation, job discovery, and professional correspondence. An MCP client can retrieve relevant LinkedIn information and support follow-up actions through the user's authenticated account.
Before deployment, operators should verify MCP client compatibility, browser authentication, session-storage handling, and whether local or HTTP transport fits their environment. They should also review which actions can change account state: messaging requires confirmation, and dismissing a job is documented as irreversible. Concurrent workloads must account for serialized browser access.
Reviews of MCP Server for LinkedIn
Be the first to share your experience with MCP Server for LinkedIn.
Write a review
Loading review form…
Alternatives to MCP Server for LinkedIn
EngramEngram is a local persistent memory system for AI coding agents, storing decisions and session summaries in SQLite and exposed through MCP and an HTTP API.
Firecrawl MCP ServerA Firecrawl connector for web search, page extraction, site mapping and bounded crawling with Markdown or structured results.