abm.dev MCP server
Account-based marketing enrichment for AI agents. Enriches people and companies from email, LinkedIn URL, or domain with cited multi-source fields.
README
<p align="center"> <img src="https://raw.githubusercontent.com/abm-dev-git/abm-mcp/main/assets/abm-diamond.svg" alt="abm.dev" width="96" /> </p>
<h1 align="center">abm.dev MCP server</h1>
<p align="center"> <strong>Account-based marketing enrichment for AI agents.</strong><br/> One endpoint. Cited, multi-source fields. Every value attributed. </p>
<p align="center"> <a href="https://www.npmjs.com/package/@abmdev/mcp"><img src="https://img.shields.io/npm/v/@abmdev/mcp?color=2563eb&label=%40abmdev%2Fmcp" alt="npm" /></a> <a href="https://github.com/abm-dev-git/abm-mcp/blob/main/LICENSE"><img src="https://img.shields.io/badge/license-MIT-green" alt="MIT" /></a> <a href="https://mcp.abm.dev/mcp"><img src="https://img.shields.io/badge/MCP-streamable--http-7c3aed" alt="MCP" /></a> <a href="https://smithery.ai/servers/abm-dev/gtm"><img src="https://smithery.ai/badge/abm-dev/gtm" alt="Smithery" /></a> </p>
- Hosted server:
https://mcp.abm.dev/mcp - Homepage: https://abm.dev
- For agents (machine-readable): https://abm.dev/llms.txt
What abm.dev is
Account-based marketing for AI agents. Give it an email, a LinkedIn URL, or a company domain; get back a structured record on the person and the company they work for. Every value carries its source, a confidence score, and a last-verified timestamp. A value is cited or it is not returned. No fabricated facts.
What you get
- Eighty-nine canonical fields per record — forty-three on the person, forty-six on the company.
- Forty signals on every prospect.
- Ten sources, resolved into one response — LinkedIn, Companies House, Perplexity, Tavily, Hunter, and others.
- Citations on every value —
source,confidence(0–1), andfreshness(last-verified timestamp).
No per-source bills. No per-field charges. No silent fallbacks.
Connect
Claude (custom connector)
- Open Settings → Connectors → Custom.
- Paste the server URL:
https://mcp.abm.dev/mcp - Complete the one-time sign-in when prompted.
Then ask, in plain language: "Enrich jane@acme.com."
Cursor / Claude Code / Windsurf
These agents drive the REST API directly. Paste this prompt:
Read https://abm.dev/llms.txt. Request an API key, save it to
.envasABM_API_KEY, then run a test enrichment againsthttps://api.abm.dev/v1/enrichments. Report the result with its citations.
Script and terminal agents
Call the REST API directly:
curl https://api.abm.dev/v1/enrichments \
-H "x-api-key: $ABM_API_KEY" \
-H "Content-Type: application/json" \
-d '{"type":"person","input":{"email":"jane@acme.com"}}'
input also accepts a LinkedIn URL or a company domain.
Install (npx)
Prefer a local command over a remote URL? The @abmdev/mcp package is a thin
bridge: npx -y @abmdev/mcp launches a local STDIO MCP server that proxies to the
hosted server at https://mcp.abm.dev/mcp, handling the OAuth 2.1 PKCE sign-in
and token caching for you. It defines no tools of its own — it exposes whatever
the hosted server serves.
Add it to any MCP client that supports command/args (Claude Desktop,
Cursor, Windsurf, Claude Code, and others):
{
"mcpServers": {
"abm": {
"command": "npx",
"args": ["-y", "@abmdev/mcp"]
}
}
}
Generic config snippet (for clients that take a bare command):
{ "command": "npx", "args": ["-y", "@abmdev/mcp"] }
On first run a browser window opens for the one-time sign-in; the token is then
cached for subsequent runs. Point the bridge at a different endpoint by setting
the ABM_MCP_URL environment variable.
Either path reaches the same hosted server. Use the custom connector (remote URL) above if your client supports it directly; use npx for clients that only speak local
command/args.
Tools
The server exposes enrichment:
- Enrich a person — from an email or a LinkedIn URL. Returns the forty-three person fields plus the company they work for.
- Enrich a company — from a domain. Returns the forty-six company fields.
Every field comes back with its source, confidence, and freshness. Real-time SSE streaming for agent loops; webhooks fire when batch jobs finish.
Pricing
Per-enrichment. No subscription. Credits never expire. All ten sources included — no add-ons.
From about €0.29 per enrichment. Packs: 30 credits €2.89, 100 €9.29, 500 €36.99 (best value), 2,000 €119.99. The playground is free. Free launch credits with code LAUNCHCODES.
See https://abm.dev/#pricing.
Links
- Homepage: https://abm.dev
- For agents (llms.txt): https://abm.dev/llms.txt
- Docs: https://abm.dev/docs · Getting started
- API reference: https://abm.dev/api-reference · Enrichment · Jobs
- For developers: https://abm.dev/for-developers · For agents: https://abm.dev/for-agents · For marketers: https://abm.dev/for-marketers
- Integrations: Claude connector · HubSpot · Notion · Slack
- Concepts: Enrichment · Confidence scores · Data sources · Canonical fields
- Sign up: https://abm.dev/sign-up · Free playground: https://abm.dev/dashboard/playground
Registry reference
This repo is the public description of a hosted MCP server. The server runs at https://mcp.abm.dev/mcp; this repo is not its source. See server.json for the machine-readable manifest.
Support
Questions or a stuck enrichment? Open a ticket: https://abm.dev
Operator
Foxley Farm Operations Ltd · UK Company No. 16392009 · Oakham, United Kingdom. GDPR-compliant.
Recommended Servers
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.
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.
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.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
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.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.