Grinfi MCP Server

Grinfi MCP Server

Connects Claude to your Grinfi.io account, enabling management of contacts, automations, messages, and more through natural language.

Category
Visit Server

README

Grinfi MCP Server

Connect Claude to your Grinfi.io account. Manage contacts, automations, messages, and more — all through natural language.


Quick Start

1. Get your API key

Go to Grinfi → Settings → API Keys and copy your key.

2. Download this repository

Click the green "Code" button above → "Download ZIP", then unzip it.

Or clone with Git:

git clone https://github.com/alex-rise/grinfi-mcp-server.git

3. Run the installer for your OS


macOS Installation

Option A: Use the installer (recommended)

  1. Open the grinfi-mcp-server folder in Finder
  2. Double-click install.sh
  3. When prompted, paste your Grinfi API key and press Enter

The installer will automatically:

  • Install Homebrew (if not installed)
  • Install Node.js (if not installed)
  • Install all dependencies and build the server
  • Configure Claude Desktop

Note: If macOS asks for your password, enter your Mac login password and press Enter. This is normal — it's installing system packages.

Option B: Manual setup (macOS)

cd grinfi-mcp-server
npm install
npm run build

Then add this to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "grinfi": {
      "command": "node",
      "args": ["/full/path/to/grinfi-mcp-server/dist/index.js"],
      "env": {
        "GRINFI_API_KEY": "your-api-key-here"
      }
    }
  }
}

Windows Installation

Option A: Use the installer (recommended)

  1. Open the grinfi-mcp-server folder in File Explorer
  2. Double-click install.bat
  3. If Windows shows a security warning ("Windows protected your PC"), click "More info" then "Run anyway"
  4. When prompted, paste your Grinfi API key and press Enter

The installer will automatically:

  • Install Node.js via winget (if not installed)
  • Install all dependencies and build the server
  • Configure Claude Desktop

Option B: Manual setup (Windows)

cd grinfi-mcp-server
npm install
npm run build

Then add this to %APPDATA%\Claude\claude_desktop_config.json:

{
  "mcpServers": {
    "grinfi": {
      "command": "node",
      "args": ["C:/full/path/to/grinfi-mcp-server/dist/index.js"],
      "env": {
        "GRINFI_API_KEY": "your-api-key-here"
      }
    }
  }
}

After Installation (All Platforms)

  1. Restart Claude Desktop — quit completely (not just close the window) and reopen
  2. Look for "grinfi" in the tools list (hammer icon at the bottom of the chat)
  3. Start chatting! Try: "Show me all my contacts"

Claude Code Configuration

claude mcp add grinfi -- node /full/path/to/grinfi-mcp-server/dist/index.js

Then set the environment variable before running Claude Code:

export GRINFI_API_KEY="your-api-key-here"
claude

Cloud Endpoint (Paid Claude plans)

If you have a paid Claude plan, you can use our hosted cloud endpoint — no local installation needed.

Visit mcp.grinfi.io to generate your personal MCP endpoint URL.


Available Tools

Contacts

Tool Description
find_contact Find a contact by LinkedIn ID, email, or name + company
search_contacts Search contacts with filters, sorting, pagination
get_contact Get contact details by UUID
update_contact Update contact fields
delete_contact Delete a contact
upsert_contact Create or update a contact in a list

Lists

Tool Description
list_lists Get all contact lists
get_list Get a list by UUID
create_list Create a new list

Automations

Tool Description
list_automations Get all automations
start_automation Start an automation
stop_automation Stop an automation
add_contact_to_automation Add existing contact to automation
add_new_contact_to_automation Create contact and add to automation
cancel_contact_from_automations Cancel contact from specific automations
cancel_contact_from_all_automations Cancel contact from all automations

Unibox (Messages)

Tool Description
list_linkedin_messages List LinkedIn messages
get_unread_conversations Get contacts with unread messages
mark_conversation_as_read Mark a LinkedIn conversation as read
send_linkedin_message Send a LinkedIn message
list_emails List emails
send_email Send an email

Sender Profiles

Tool Description
list_sender_profiles List all sender profiles
get_sender_profile Get sender profile details
create_sender_profile Create a new sender profile

CSV Import / Export (new in 1.1)

Tool Description
upload_csv Upload a CSV file from local disk; returns a file_import UUID
import_leads_from_file Import contacts from an uploaded CSV with column mapping
import_companies_from_file Import companies from an uploaded CSV
export_leads_csv Queue a CSV export of contacts matching a filter
export_companies_csv Queue a CSV export of companies matching a filter
download_export Get the download payload for a queued export

Lead Enrichment & Analytics (new in 1.1)

Tool Description
enrich_leads Trigger advanced LinkedIn enrichment for contacts
count_leads Count contacts matching a filter
get_leads_metrics Get team engagement metrics for contacts

Automation Folders & Enrollment

Tool Description
list_automation_folders List automation folders (renamed from list_flow_workspaces in 1.2)
create_automation_folder Create a new automation folder (renamed in 1.2)
update_automation_folder Rename or change display order (renamed in 1.2)
delete_automation_folder Delete a folder (automations inside are not deleted) (renamed in 1.2)
list_flow_leads Search contacts enrolled in automations
delete_flow_lead_history Delete a contact's automation history (full erase)

AI Agents & Templates — full CRUD (new in 1.1)

Tool Description
create_ai_agent Create a new AI agent
update_ai_agent Update an AI agent
delete_ai_agent Delete an AI agent
update_ai_template Update an AI template
delete_ai_template Delete an AI template

Closures (new in 1.1)

Tool Description
upload_attachment Upload a local file as an attachment
remove_from_leads_blacklist Remove a contact from the leads blacklist
remove_from_companies_blacklist Remove a company from the companies blacklist
update_custom_field Rename a custom field or change order
delete_custom_field Delete a custom field (and all its values)

Failed-task triage (new in 1.2)

Tool Description
diagnose_failed_tasks Read-only triage: failure breakdown by error code, per-sender, sample tasks, pattern hints (captcha/proxy/rate-limit detection)
restart_failed_tasks Bulk-retry failed automation tasks (filter or explicit UUIDs); supports dry_run
skip_failed_tasks Bulk-skip failed tasks (lead progresses to next node)
restart_failed_tasks_from_top Re-enrol matched leads from the start of a flow; optionally swap to a different sender

Workspace dashboard & health (new in 1.2)

Tool Description
workspace_health_check Single-call 6-section operational overview: today/yesterday counts, LinkedIn fleet, mailbox health, active flows, failed tasks
get_dashboard CRM analytics widgets: activities-over-time, conversion-funnel, pipeline-distribution, sender-performance, response-rate
send_volume_report Period-aware outreach event counts with flexible group_by (type, status, flow, sender, day, week)
diagnose_mailbox Deep mailbox health diagnosis with detected issues + recommendations
get_health_snapshots Fleet-wide LinkedIn account snapshot (status, cookie, health score, daily caps)

LinkedIn browsers (new in 1.2)

Tool Description
list_linkedin_browsers List all LinkedIn browser profiles
get_linkedin_browser Get a browser by ID
create_linkedin_browser Create a browser linked to a sender profile
delete_linkedin_browser Delete a browser (irreversible)
run_linkedin_browser / stop_linkedin_browser Start / stop a browser session
set_linkedin_browser_proxy Change proxy country
share_linkedin_browser Share browser access with team members by email
diagnose_linkedin_browser Deep health diagnosis on one browser (cookie, health score, limits, proxy)

Data sources (LinkedIn import jobs) (new in 1.2)

Tool Description
list_data_sources / get_data_source Discover & inspect import jobs
create_data_source Create import from CSV / Sales Navigator / LinkedIn search / post engagement
update_data_source / delete_data_source Edit or remove an import job

Account & teams (new in 1.2)

Tool Description
get_current_user Authenticated user profile
list_teams All teams accessible via the Grinfi API (vs list_my_teams for locally-configured keys)
get_team Team details by numeric ID

Integrations & diagnostics (new in 1.2)

Tool Description
call_external_api Ad-hoc outbound HTTP request (SSRF-protected, public URLs only, 1MB cap)
list_outbound_log Outbound HTTP log: webhook deliveries, automation API calls, enrichment requests
test_llm_connection Smoke-test a stored LLM integration (~5 token cost)

Example Conversations

Find a contact:

"Find the contact with email john@example.com"

Search with filters:

"Show me contacts at Google created after January 2024"

Manage automations:

"List my automations and start the one called CMO Germany"

Send messages:

"Send a LinkedIn message to Anna saying we have a new case study"


Troubleshooting

"GRINFI_API_KEY is not set"

Make sure your API key is configured. Re-run the installer or edit the Claude Desktop config file manually (see Manual Setup sections above).

Tools not showing in Claude

  1. Make sure Claude Desktop is fully restarted (quit and reopen, not just close window)
  2. Check the config file path is correct for your OS
  3. Make sure the path to dist/index.js is absolute (starts with / on Mac or C:\ on Windows)

API errors (401/403)

Your API key may be invalid or expired. Generate a new one in Grinfi → Settings → API Keys.

Build errors

cd grinfi-mcp-server
rm -rf node_modules dist
npm install
npm run build

Installer won't start (Mac)

If double-click doesn't work, open Terminal and run:

bash /Users/yourname/Documents/grinfi-mcp-server/install.sh

Installer won't start (Windows)

Make sure you're double-clicking install.bat (not install.ps1). If Windows blocks it, click "More info" → "Run anyway".

If that doesn't work, open PowerShell manually (press Win + X, choose "PowerShell"), then run:

cd "$HOME\Documents\grinfi-mcp-server"
powershell -ExecutionPolicy Bypass -File install.ps1

Requirements


License

MIT

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured