Xverum

# Xverum — MCP Server

Install in Cursor Install in VS Code

[![License: MIT](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE) ![MCP Compatible](https://img.shields.io/badge/MCP-Compatible-blueviolet) ![Claude Ready](https://img.shields.io/badge/Claude-Ready-orange) ![OpenAI Compatible](https://img.shields.io/badge/OpenAI-Compatible-lightgrey) ![GDPR Compliant](https://img.shields.io/badge/GDPR-Compliant-green) ![CCPA Compliant](https://img.shields.io/badge/CCPA-Compliant-green) Find and enrich people from 750M professional profiles, with Next Move Signal job-change scores. Search by role, seniority, skills, industry, and location in plain English. Pull profiles with description, work history, education, and seniority, then see who's likely to change jobs next with Next Move Signal. Built for AI products and agents: sourcing candidates, building lead lists, and mapping markets and accounts. This is a **hosted, remote MCP server** — there is nothing to install or run. Point your client at the endpoint below and sign in with OAuth. ## Is there an MCP server for finding people? Yes. The Xverum MCP lets your AI assistant find and enrich the right people from a global professional-profiles dataset — no scraping, no browser tabs. Ask in plain language (*"find senior data engineers in Amsterdam"* or *"VPs of sales at Series A SaaS companies"*) and it returns ranked profiles; ask about one and it pulls the full background. It's built for two jobs: sourcing job candidates by role, skills, and location, and building sales lead lists of decision-makers. Connect it to Claude or any MCP-compatible assistant in one command. ## Prerequisites Sign up at [Ask Xverum](https://ask.xverum.com/). ## Connect your client ```bash claude mcp add --transport http xverum https://mcp.xverum.com/mcp ```
Claude Code — verify Verify with `/mcp` — you should see `xverum` with five tools.
Cursor Add to `~/.cursor/mcp.json`: ```json { "mcpServers": { "xverum": { "type": "http", "url": "https://mcp.xverum.com/mcp" } } } ```
Visual Studio Code Add to `.vscode/mcp.json`: ```json { "servers": { "xverum": { "type": "http", "url": "https://mcp.xverum.com/mcp" } } } ```
Any other MCP client Most clients take a JSON config of this shape: ```json { "mcpServers": { "xverum": { "type": "http", "url": "https://mcp.xverum.com/mcp" } } } ``` Consult your client's docs for where that file lives.
Also in the MCP Registry as `com.xverum/mcp`, so many agents can add it by name. > **Authentication.** The server uses OAuth — on first connection your client opens a > browser window to sign in. No API keys to paste. See [docs/authentication.md](docs/authentication.md). ## Tools | Tool | What it does | Cost | |------|--------------|------| | `search_people_xverum` | Find people matching a natural-language description – candidates, prospects, or decision-makers. Returns ranked profiles. | 1 credit per result | | `enrich_person_xverum` | Pull the full profile for one person by id: description, work history, education, and seniority. | 4 credits per full profile | | `predict_job_change_xverum` | Score how likely a person is to change roles – before they declare it. Refreshed weekly. | 10 credits per score | | `search_company_xverum` | Find companies from a natural-language description. Returns a ranked page of company summary cards. | 1 credit per result | | `enrich_company_xverum` | Pull the full record for one company. | 4 credits per full company | Full reference: [docs/tools.md](docs/tools.md). ### Examples **Example 1: Search people** (`search_people_xverum`) ``` Senior Python developers with Kubernetes in Seattle ``` **Example 2: Enrich a profile** (`enrich_person_xverum`) ``` Get profile id 401241903 ``` **Example 3: Predict job changes** (`predict_job_change_xverum`) ``` Use predict_job_change_xverum for profile ID 271961733. ``` **Example 4: Search companies** (`search_company_xverum`) ``` cybersecurity companies in Germany ``` ``` nonprofit organizations working on climate change in California ``` **Example 5: Enrich a company** (`enrich_company_xverum`) ``` Get company id 88673718 ``` ## Credits 1 credit per search result · 4 credits per full profile · 10 credits per job-change score · 1 credit per company search result · 4 credits per full company. Each response reports `credits_used` and `credits_remaining`. [See plans and pricing](https://ask.xverum.com/pricing) ## Docs - [Quickstart](docs/quickstart.md) — connect and run your first search. - [Tools](docs/tools.md) — parameters, return shapes, and worked examples. - [Authentication](docs/authentication.md) — OAuth sign-in and what we store. - [Troubleshooting](docs/troubleshooting.md) — error codes and what to do about them. - [Versioning](docs/versioning.md) — versioning scheme and deprecation policy. ## Privacy & security - Authentication is handled via OAuth — no API keys are stored or transmitted in config files. - The server exposes exactly the five tools listed above. It cannot write, delete, or take any action on your behalf. - Profile data comes from Xverum's licensed professional-profiles dataset. ## Links - [Ask Xverum](https://ask.xverum.com/docs/mcp-server/) — website - [MCP Registry](https://registry.modelcontextprotocol.io) — `com.xverum/mcp` ## Support Questions or issues: open an issue on this repo, or contact us via [agentic.support@xverum.com](mailto:agentic.support@xverum.com). [![smithery badge](https://smithery.ai/badge/xverum/xverum-mcp)](https://smithery.ai/servers/xverum/xverum-mcp)