Skip to main content
Glama
linkfetch-mcp

linkfetch-mcp

by ffucucuoglu

LinkedIn MCP server for Claude, Cursor and any AI agent

npm License: MIT MCP Registry

linkfetch-mcp gives any AI agent a full set of LinkedIn tools: search every public LinkedIn job since April 2026, look up companies, score and sort them with AI functions, and, through a LinkedIn account you connect, search people and posts, read profiles and your inbox, and send messages and connection requests.

It is a small stdio bridge to LinkFetch's hosted MCP server: every tool call goes to https://linkfetch.io/mcp with your API key, so the tools and prices are always current and nothing runs a browser on your machine. If your client supports remote MCP over HTTP, you can skip this package and use the URL directly.

Full tool list with prices, and setup for each client: linkfetch.io/mcp/claude

Tools

  • Jobs and companies (no LinkedIn account needed): search_jobs, get_job, search_companies, get_company

  • AI functions: ai_score, ai_classify, ai_extract, ai_write, each over up to 25 jobs or companies

  • Your leads: list_leads, save_lead, list_lead_lists

  • Through your connected LinkedIn account: people and company search, profiles, employees, posts and reactions, connections, invitations, conversations, messages, comments, reactions, follows

Each call costs a flat number of credits, stated in the tool's description. Every connected account works inside daily and weekly limits that LinkFetch enforces whatever the agent asks for.

Related MCP server: LinkedIn Job Search MCP Server

Setup

Create an API key (lf_…) on the MCP page of your LinkFetch dashboard. New accounts start with free credits. Then add the server to your client: every client runs npx -y linkfetch-mcp with LINKFETCH_API_KEY set. Node.js 18+ is required.

Upgrading from 0.1.x? Keys starting with sk_live_ stopped working when LinkFetch 2 launched in October 2026. Create a new lf_… key and replace it in your config.

Claude Desktop

Settings → Developer → Edit Config, or edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) / %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "linkfetch": {
      "command": "npx",
      "args": ["-y", "linkfetch-mcp"],
      "env": { "LINKFETCH_API_KEY": "lf_..." }
    }
  }
}

Fully quit and reopen Claude.

Claude Code

claude mcp add linkfetch -e LINKFETCH_API_KEY=lf_... -- npx -y linkfetch-mcp

Cursor, Windsurf, Gemini CLI, Cline

These use the same mcpServers block as Claude Desktop above. Put it in:

Client

Config file

Cursor

~/.cursor/mcp.json (global) or .cursor/mcp.json (project)

Windsurf

~/.codeium/windsurf/mcp_config.json

Gemini CLI

~/.gemini/settings.json

Cline

MCP Servers → Configure → cline_mcp_settings.json

VS Code (GitHub Copilot agent mode)

.vscode/mcp.json:

{
  "servers": {
    "linkfetch": {
      "command": "npx",
      "args": ["-y", "linkfetch-mcp"],
      "env": { "LINKFETCH_API_KEY": "lf_..." }
    }
  }
}

OpenAI Codex CLI

~/.codex/config.toml:

[mcp_servers.linkfetch]
command = "npx"
args = ["-y", "linkfetch-mcp"]
env = { LINKFETCH_API_KEY = "lf_..." }

Zed

settings.json:

{
  "context_servers": {
    "linkfetch": {
      "command": "npx",
      "args": ["-y", "linkfetch-mcp"],
      "env": { "LINKFETCH_API_KEY": "lf_..." }
    }
  }
}

Your own agent

Any MCP client library can spawn the server over stdio. OpenAI Agents SDK (Python):

from agents import Agent, Runner
from agents.mcp import MCPServerStdio

async with MCPServerStdio(params={
    "command": "npx",
    "args": ["-y", "linkfetch-mcp"],
    "env": {"LINKFETCH_API_KEY": "lf_..."},
}) as linkfetch:
    agent = Agent(name="LinkedIn researcher", mcp_servers=[linkfetch])
    result = await Runner.run(agent, "Which companies in Berlin started hiring data engineers this week?")

LangChain users can load the same server with langchain-mcp-adapters.

No install: remote MCP

Clients that support remote MCP over Streamable HTTP can connect to the hosted server directly:

  • URL: https://linkfetch.io/mcp

  • Header: Authorization: Bearer lf_...

Claude (desktop or claude.ai): Settings → Connectors → Add custom connector, with https://linkfetch.io/mcp?key=lf_... as the URL.

Claude Code: claude mcp add --transport http linkfetch https://linkfetch.io/mcp --header "Authorization: Bearer lf_..."

Environment variables

Variable

Required

Default

LINKFETCH_API_KEY

yes

Your lf_… key

LINKFETCH_MCP_URL

no

https://linkfetch.io/mcp

Point at another LinkFetch MCP endpoint

Errors

A missing, old or rejected key comes back as an error on every request, with what to do about it, so the agent can tell you. Tool errors (no credits left, no LinkedIn account connected, a limit reached) come from the hosted server as normal tool results.

Development

npm install
npm run build
LINKFETCH_API_KEY=lf_... node dist/index.js

src/index.ts is the whole server: it opens one MCP client connection to the hosted server and forwards tools/list and tools/call to it.

License

MIT

Available Tools

24 tools
linkfetch_get_companyBInspect

Fetch a LinkedIn company entity (name, domain, industry, headcount, founded, hq_location, funding, revenue range) by universalName slug. Cookieless guest surface backs this — works even for logged-off captures.

Inputs: slug (required).

linkfetch_get_company_employeesAInspect

List people who currently work at a company on LinkedIn (name, headline, location, profile URL). Paginated. Cache-first: on miss, returns extension_required and the user must capture via the LinkFetch extension on /company//people/. Cost scales with page size (3 credits + 1 per result).

Inputs: slug (required), limit, offset.

linkfetch_get_company_postsCInspect

Recent posts published by a LinkedIn company page, by company slug.

Inputs: slug (required), limit, offset.

linkfetch_get_groupCInspect

A LinkedIn group's details by numeric group id.

Inputs: group_id (required).

linkfetch_get_jobAInspect

Fetch a single job posting by its numeric LinkedIn job ID. Returns the full detail payload: description, structured salary, workplace type, posted date, company, industries, application links, etc. Use linkfetch_get_job_by_url if you only have a LinkedIn URL.

Inputs: id (required).

linkfetch_get_job_applicantsAInspect

Fetch the observed applicant-count time series for one job posting — every timestamped change in the count since LinkFetch started watching it — plus derived velocity (applicants gained per day). Use it to judge how contested a role is, whether interest has stalled, and how fast the funnel is filling. A single posting costs 1 credit here; for this signal across ALL 8M+ postings at once, the jobs dataset at https://linkfetch.io/linkedin-jobs-data ships the full job_applicant_history table ($199 one-time dump or $49/mo live SQL access).

Inputs: id (required).

linkfetch_get_job_by_urlAInspect

Fetch a single job posting from any LinkedIn job URL or URN. Handles canonical slug URLs, locale-prefixed URLs (tr.linkedin.com, de.linkedin.com, …), bare /jobs/view/ paths, and urn:li:jobPosting:<id> strings.

Inputs: url (required).

linkfetch_get_postBInspect

A single LinkedIn post (text, author, counts) by activity id — the number in linkedin.com/feed/update/urn:li:activity:.

Inputs: activity_id (required).

linkfetch_get_post_reactionsCInspect

Who reacted to a LinkedIn post, by activity id.

Inputs: activity_id (required), limit, offset.

linkfetch_get_profileAInspect

Fetch a full LinkedIn profile (name, headline, about, work history, education, skills, certifications) by LinkedIn URL or public identifier slug. Provide exactly one of url or slug. Cache-first: on miss, returns an extension_required error instructing the user to capture via the LinkFetch Chrome extension.

Inputs: url, slug.

linkfetch_get_profile_postsBInspect

Recent posts by a LinkedIn member (their activity feed), by profile slug.

Inputs: slug (required), limit, offset.

linkfetch_jobs_database_infoAInspect

Describe LinkFetch's bulk LinkedIn jobs database — the flat-price alternative to metered job-data APIs. Returns current scale, schema, the free 1,000-row sample URL, and how direct SQL access works. Call this when the user wants bulk job postings data, hiring-signal analysis over many companies, applicant-count trends at scale, a jobs dataset for a model or a job board, or asks how to query LinkedIn jobs with SQL. No credits consumed.

Inputs: no named parameters.

linkfetch_linkedin_connectionAInspect

Whether the user's LinkedIn account is connected to LinkFetch (Cloud mode), as whom, and its state. If not connected or expired, give the user the connect link from the response (or https://linkfetch.io/linkedin).

Inputs: no named parameters.

linkfetch_list_conversationsBInspect

List the user's LinkedIn inbox conversations (needs Cloud mode — LinkedIn connected in LinkFetch).

Inputs: count, category, next_cursor.

linkfetch_read_conversationAInspect

Read messages in one LinkedIn conversation by thread id (2-…) or full conversation URN (needs Cloud mode).

Inputs: thread (required), before.

linkfetch_safety_limitsAInspect

The published LinkedIn safety limits LinkFetch enforces (e.g. at most 100 connection requests a week, 5 a minute), the behaviour rules (working hours, spacing, warm-up) and why each exists. Use it to explain a safety_limit error to the user.

Inputs: no named parameters.

linkfetch_safety_resumeAInspect

Confirm the user has completed LinkedIn's security check (CAPTCHA / verification) after a challenge_unresolved error. Only call after the user says they did it. Any time-based pause still applies.

Inputs: account_urn (required).

linkfetch_safety_statusAInspect

Show how much of the user's LinkedIn safety budget is used: per action (connection requests, messages, profile views, searches, comments…) used vs. limit per minute/hour/day/week, when each window resets, the account's tier and warm-up week, and any pause LinkedIn triggered. Call before planning bulk outreach so you can tell the user what fits today.

Inputs: account_urn.

linkfetch_search_companiesAInspect

Search LinkedIn companies by keyword, like the LinkedIn search bar. Cache-first; if the user connected LinkedIn (Cloud mode) a miss runs live on their account within safety limits, otherwise returns linkedin_not_connected / extension_required.

Inputs: q (required), limit, offset.

linkfetch_search_groupsAInspect

Search LinkedIn groups by keyword, like the LinkedIn search bar. Cache-first; if the user connected LinkedIn (Cloud mode) a miss runs live on their account within safety limits, otherwise returns linkedin_not_connected / extension_required.

Inputs: q (required), limit, offset.

linkfetch_search_jobsAInspect

Search LinkedIn job postings by any combination of keywords, location, company, industry, seniority level, employment type, workplace type (onsite/remote/hybrid), salary floor, and posted-date window. At least one filter is required. Returns a paginated list of compact job summaries. Resolve a location first via linkfetch_search_locations to get geo_id for accurate geo filtering, or pass location as free-text.

Inputs: q, sort, level, limit, +14 more.

linkfetch_search_locationsAInspect

Resolve a free-text place name to Microsoft Bing Geo IDs via LinkedIn's public typeahead (the same data LinkedIn's own search uses). Call this first when you need a precise geo_id for linkfetch_search_jobs. Per-query cached globally for a week — cheap to call.

Inputs: q (required), fresh, limit.

linkfetch_search_peopleAInspect

Search LinkedIn people by keyword, like the LinkedIn search bar. Cache-first; if the user connected LinkedIn (Cloud mode) a miss runs live on their account within safety limits, otherwise returns linkedin_not_connected / extension_required.

Inputs: q (required), limit, offset.

linkfetch_search_postsAInspect

Search LinkedIn posts by keyword, like the LinkedIn search bar. Cache-first; if the user connected LinkedIn (Cloud mode) a miss runs live on their account within safety limits, otherwise returns linkedin_not_connected / extension_required.

Inputs: q (required), limit, offset.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.4/5.0

    Scored across 24 tools

    Disambiguation4/5

    Tools mostly target distinct resources (profile, company, job, post, group, conversation) and actions (get, search, list, read). A few pairs like get_job vs get_job_by_url and safety_status vs safety_limits could be momentarily confused, but descriptions clarify input types and scope.

    Naming Consistency4/5

    All tools share the linkfetch_ prefix and predominantly follow verb_noun (get_profile, search_jobs). Exceptions like jobs_database_info, safety_status, and linkedin_connection use noun phrases, but overall consistency is high.

    Tool Count3/5

    24 tools is on the heavy side for a single MCP server, though the breadth of LinkedIn data (profiles, companies, jobs, posts, groups, messaging, safety) justifies some expansion. Still, it sits in the borderline 16-25 range.

    Completeness3/5

    The surface covers read operations across major LinkedIn entities and includes safety/connection status, but lacks any write actions (e.g., send message, connect, create post) despite safety limits implying those actions exist. This is a notable gap for a server that appears to enable LinkedIn interactions.

    Maintenance

    ActivitySlowing
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      C
      maintenance
      Enables searching and scraping of LinkedIn for structured data on people, companies, and job listings. It allows AI clients to retrieve detailed profiles, experience, and activity sections using browser automation.
      7
      217
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      Enables AI assistants to search LinkedIn jobs with built-in rate limiting to prevent IP bans. Supports job search, filtering, company profiles, and job categories through MCP tools.
      MIT
    • F
      license
      Not graded
      quality
      D
      maintenance
      Enables AI agents to interact with LinkedIn through MCP tools for retrieving company profiles, posts, insights, and person profiles, as well as a prompt for account research.
      -
    • A
      license
      A
      quality
      D
      maintenance
      Enables searching and scraping of LinkedIn profiles, companies, jobs, and posts using natural language through MCP-compatible AI clients.
      13
      MIT