I built an open MCP server for German employment law, every number with its source

Description and use

An open MCP server for German employment law delivers verified numeric answers with sources, reducing compliance risk for HR professionals.

Problem

HR professionals need accurate, up‑to‑date numeric answers for German employment law but LLMs provide unverified numbers.

Outcome

The MCP server supplies precise, source‑verified numeric answers for German employment law, reducing compliance risk and freeing HR staff from manual calculations.

Workflow

• 1. Deploy the MCP server (cloud, Docker, local).
• 2. Connect the assistant (Claude, ChatGPT, Gemini) to the server via custom connector.
• 3. Ask a legal question in natural language.
• 4. Assistant calls the relevant MCP tool (e.g., kuendigungsfrist_rechnen).
• 5. Tool returns structured JSON with result, calculation steps, sources, and disclaimer.
• 6. Assistant presents answer to user.
• 7. If value not confirmed, tool indicates status and check date.

Evidence

People now ask these questions in Claude or ChatGPT. The text is usually fine. The numbers are the problem: a date or an amount that comes from model memory, with no way to check it.

Source

An MCP server is a set of tools an assistant can call instead of answering from memory.

Source

Every answer has the same shape: ergebnis, rechenweg, quellen, fallen, naechster_schritt and hinweis.

Source

The server only calculates. It never calls a model. Your assistant writes the text, the numbers come from the tool.

Source

The remote server is https://hr.lintelly.com/mcp (Streamable HTTP). There is no account, no key and no sign-in.

Source