Model Context Protocol

Control Updately from your AI

38 AI-callable tools that let Claude, Gemini, Cursor, and any MCP-compatible client manage your LinkedIn outreach — campaigns, contacts, signals, social listening, lead search, and ICP via natural language.

$npm i @updately/mcp-server
38
AI-Callable Tools
12
Tool Categories
v1.0.3
Latest Version
< 5min
Setup Time
How it works

Your AI talks to Updately, natively

The MCP protocol creates a direct bridge between your AI assistant and Updately's full API — no code needed.

Step 1

You speak

"Create a campaign targeting VP of Engineering at Series B SaaS companies"

Step 2

MCP translates

Your AI calls create_campaign, create_icp, and enroll_contacts tools automatically

Step 3

Updately executes

Campaign is created, ICP is set, contacts are enrolled — all in seconds

Use Cases

Just ask in natural language

The MCP server translates your intent into API calls — no JSON, no endpoints, no documentation.

🚀

"Create a new outreach campaign targeting DevOps engineers"

Creates campaign with AI-optimized messaging templates

💬

"Show me all contacts who replied in the last 7 days"

Filters and displays engaged leads with reply details

📨

"Reach out to this LinkedIn profile with a connection invite"

Auto-detects status, sends DM or invite with personalized note

📊

"Pause all active campaigns and show me the stats"

Pauses campaigns and returns conversion metrics

📡

"Create a signal for LinkedIn job posts mentioning AI"

Sets up real-time monitoring for high-intent hiring signals

🎯

"Define my ICP as Series A SaaS companies with 50-200 employees"

Creates an ICP profile for automatic lead scoring

🔍

"Search for DevOps engineers at startups in San Francisco"

Finds matching LinkedIn profiles with name, headline, company & URL

📡

"Fetch recent Reddit and Twitter posts about social listening tools"

Returns scored posts with sentiment, priority, and lead-type tags

Quick Setup

Get up and running in under 5 minutes

Install, configure, and start managing outreach from your AI assistant.

01

Install the package

Install from npm and build.

npm i @updately/mcp-server
npm run build
02

Generate an API Key

Create an API key from your Updately dashboard or via the API.

curl -X POST https://api.updately.ai/org/api-key/generate \
  -H "Authorization: Bearer YOUR_ORG_ID" \
  -H "Content-Type: application/json" \
  -d '{"name": "MCP Server"}'
03

Configure your AI Tool

Add the server to your Claude Desktop, Cursor, or any MCP-compatible client.

{
  "mcpServers": {
    "updately": {
      "command": "node",
      "args": ["/path/to/updately-mcp-server/build/index.js"],
      "env": {
        "UPDATELY_API_KEY": "api_key_your-key-here",
        "UPDATELY_API_URL": "https://api.updately.ai"
      }
    }
  }
}
04

Start using natural language

Ask your AI assistant to manage campaigns, contacts, signals, and more — in plain English.

API Surface

38 Available Tools

Full control over your LinkedIn outreach pipeline — campaigns, contacts, signals, lead search, social listening, ICP, inbox, and more.

C
Confirm Lead Action
Lead Search
Create Campaign
Campaigns
Create Icp
ICP (Ideal Customer Profile)
Create Lead List
Lead Lists
Create Signal
Signals
D
Delete Campaign
Campaigns
Delete Icp
ICP (Ideal Customer Profile)
Delete Signal
Signals
Delete Step
Workflow Steps
E
Enroll Contacts
Contacts
F
Fetch Posts
Social Listening
G
Get Account Usage
LinkedIn Accounts
Get Campaign
Campaigns
Get Campaign Stats
Campaigns
Get Contact Messages
Inbox
Get Leads In List
Lead Lists
Get Saved Posts
Social Listening
Get Signal
Signals
L
List Accounts
LinkedIn Accounts
List Campaigns
Campaigns
List Contacts
Contacts
List Icps
ICP (Ideal Customer Profile)
List Keyword Groups
Social Listening
List Lead Lists
Lead Lists
List Signals
Signals
List Steps
Workflow Steps
P
Process Contact
Scheduler
R
Reach Out
Direct Outreach
Remove Contact
Contacts
Run Campaign
Scheduler
Run Signal
Signals
S
Save Post
Social Listening
Search Leads
Lead Search
Send Message
Inbox
U
Update Campaign Status
Campaigns
Update Icp
ICP (Ideal Customer Profile)
Update Signal
Signals
Upsert Step
Workflow Steps
Compatibility

Works with your favorite tools

Any MCP-compatible AI client works out of the box.

Claude Desktop

By Anthropic

Cursor

AI Code Editor

Windsurf

By Codeium

Any MCP Client

Open Protocol

Developer Experience

Batteries included

Built with TypeScript, fully typed, and ready to use. The MCP server handles authentication, error handling, pagination, and rate limiting — so your AI can focus on the task at hand.

TypeScript with full type definitions
Automatic error handling & retries
MCP Inspector for local testing
Watch mode for development
Published on npm — always up to date
terminal — npm
$ npm i @updately/mcp-server
added 12 packages in 3.2s
$ npm run build
✓ Compiled successfully
$ npm run dev
⚡ MCP Server running — watching for changes
38 tools registered • Ready for connections
$ npm run inspect
🔍 MCP Inspector → http://localhost:5173

Ready to automate with AI?

Install the MCP server, generate your API key, and start managing LinkedIn outreach from your AI assistant in minutes.