Server data from the Official MCP Registry
Cross-border pension facts for 41 countries: coverage, sourced rules and dated deadlines.
About
Cross-border pension facts for 41 countries: coverage, sourced rules and dated deadlines.
Remote endpoints: streamable-http: https://mcp.pensionhunter.ai
Security Report
Valid MCP server (1 strong, 1 medium validity signals). No known CVEs in dependencies. Imported from the Official MCP Registry.
6 tools verified · Open access · No issues found
Security scores are indicators to help you make informed decisions, not guarantees. Always review permissions before connecting any MCP server.
Permissions Found in Source Code
Found by scanning the linked source code. This listing connects to a hosted endpoint, so none of this runs on your machine: it describes what the server software does where it is hosted.
How to Connect
Remote Plugin
No local installation needed. Your AI client connects to the remote endpoint directly.
Add this to your MCP configuration to connect:
{
"mcpServers": {
"ai-pensionhunter-pensionhunter": {
"url": "https://mcp.pensionhunter.ai"
}
}
}Documentation
View on GitHubFrom the project's GitHub README.
PensionHunter MCP server
PensionHunter identifies pensions people may have left in countries they worked in. Our MCP server is remote and read-only (the one exception, create_order_link, opens a checkout only when the person asks to order), needs no account and no sign-in, and runs nothing on your computer. It is served over streamable HTTP at https://mcp.pensionhunter.ai and is listed in the MCP Registry as ai.pensionhunter/pensionhunter. Any MCP client that supports remote HTTP servers can use it. Every fact it returns carries its source and the date we read it.
Tools
list_covered_countrieslists the countries we cover, each with one sourced line on what exists there.country_pension_factsreturns the stored facts for one country: what the money is worth, why the free route fails from abroad, and dated obligations. For a country we have mapped but not yet researched in depth, it returns the paying institution and its official route instead, markedtier: mapped, with no offer.check_work_historytakes a list of work periods, each a country and years, and says for each country whether a pension is likely to exist and why.resolve_employerreads the country an employer name implies, from evidence in the name itself. It also says what the employer is called today, read by exact name from official registers such as the Global LEI Index (GLEIF), with the register and the date we read it, and it names the employer's country register of occupational pension schemes or says that the country has none.authority_answertakes a country and a question in plain words and returns what a national pension institution told us in writing, as published on our answers page: the institution, its answer exactly as published, our translation where there is one, the date of the reply and the link to that answer on https://pensionhunter.ai/answers. When nothing matches it lists the questions we hold answers to for that country. When we have written to the country and hold no answer yet, it says so with the date we asked. It names institutions only, never a person.
And one tool that is not read-only:
create_order_linkopens a secure Stripe checkout for one of our reports, for the countries the person chose, and returns its link, which works for about 24 hours. Your assistant should call it only when the person asks to order. It sends only the country codes and the report tier; Stripe asks for the email address at checkout. If a checkout cannot be opened, it returns our order page with the same choices filled in.
Answers in 24 languages. Every tool takes an optional language (an ISO 639-1 code such as fr, de, ja or pt-BR). Our own sentences then come back in that language from committed tables, marked as translations of our own wording with review recommended. Facts, numbers, dates and authority answers stay exactly as stored, in English or the language of the source.
Use in VS Code / GitHub Copilot
Add this to .vscode/mcp.json in your workspace, or to your user profile configuration (run MCP: Open User Configuration):
{
"servers": {
"pensionhunter": {
"type": "http",
"url": "https://mcp.pensionhunter.ai"
}
}
}
You can also run MCP: Add Server, choose HTTP and enter https://mcp.pensionhunter.ai. The tools then appear in Copilot Chat in agent mode.
Other clients (Cursor, Windsurf and the like): add a remote HTTP MCP server named pensionhunter with the URL https://mcp.pensionhunter.ai. No headers or keys are needed.
Use in Claude
This repository is also a Claude plugin. It adds the same server, together with the pensionhunter skill, which tells Claude when the server fits, which tool to call, what to send, and how to present results: every fact with its source and read date, no financial advice, no personal amounts, the order link offered once, exactly as the tool returns it, and create_order_link only when the person asks to order.
In claude.ai, the Claude desktop app or Cowork. Open Customize, then Plugins, find PensionHunter in Discover, and select Add. Then open the plugin's Connectors tab and add the PensionHunter connector. It needs no sign-in.
In Claude Code. A plugin added on claude.ai syncs to Claude Code when you sign in with the same account; run /reload-plugins to load it. If you have a copy of this plugin folder, you can also load it for one session:
claude --plugin-dir ./claude-plugin
The connector on its own. You can also add only the connector, without the skill:
claude mcp add --transport http pensionhunter https://mcp.pensionhunter.ai
Example prompts
- "Which countries does PensionHunter cover?"
- "I worked in the Netherlands from 2004 to 2011 and in Hong Kong from 2012 to 2016. In which of those countries might a pension exist for me?"
- "What should someone who has left the Netherlands know about the Dutch state pension?"
- "What is my former employer Philips Electronics called today, and which register lists occupational pension schemes in its country?"
- "Has Ireland's national pension institution answered PensionHunter's questions in writing? Show me what it said."
Privacy
We store no personal data. The tools receive only the inputs they need: country names or codes, years, employer names, a question in plain words and, optionally, a birth year and whether the person has died. Do not send names, email addresses or other identifiers. To say what an employer is called today, our server looks the employer name up in official registers, starting with the Global LEI Index; no other input goes to a third party. Links we return carry country codes only, never any personal detail. When the person asks to order, create_order_link sends the country codes and the report tier; we log those codes and the time, nothing more, and the person enters their email address and pays on Stripe's page.
Good to know
We identify pensions and raise awareness of them. We do not move money or act for anyone, and we do not compute personal amounts. Not regulated by the FCA. We do not provide financial advice.
Support and links
- Support: hello@pensionhunter.ai
- Connector setup and details: https://pensionhunter.ai/mcp
- Privacy policy: https://pensionhunter.ai/privacy
- Terms: https://pensionhunter.ai/terms
The PensionHunter Team
Reviews
No reviews yet
Be the first to review this server!
More Developer Tools MCP Servers
Fetch
Freeby Modelcontextprotocol · Developer Tools
Web content fetching and conversion for efficient LLM usage
Git
Freeby Modelcontextprotocol · Developer Tools
Read, search, and manipulate Git repositories programmatically
Worldmonitor
Freeby Koala73 · Developer Tools
Live markets, conflicts, country risk, chokepoints, energy, and China decision signals. 93 tools.
Toleno
Freeby Toleno · Developer Tools
Toleno Network MCP Server — Manage your Toleno mining account with Claude AI using natural language.
MCP Marketplace
Freeby mcp-marketplace · Developer Tools
Search and install MCP servers from inside your AI client.
mcp-creator-python
Freeby mcp-marketplace · Developer Tools
Create, build, and publish Python MCP servers to PyPI — conversationally.
