Sloneek MCP server: How to set it up and connect your ChatGPT or other AI tools

Sloneek is becoming easier to use than ever before. With the new Sloneek MCP server, you can connect Sloneek to ChatGPT and other MCP-compatible assistants — and interact with your HR data using natural language.

Instead of clicking through multiple screens, filters, and reports, you can simply ask:

“Whose contract is about to finish next month?”
“Show me all pending absence requests.”
“Which documents are waiting for signature?”
“Update Peter’s contract end date by one year.”

The assistant understands your request, connects to Sloneek, and helps you read, analyse, or update the relevant data directly.

What is MCP?

MCP, or Model Context Protocol, is a standard that allows AI tools to connect securely to external systems and work with their data.

In simple terms, MCP works as a bridge between your AI assistant and Sloneek.

That means your assistant can access Sloneek data and perform actions based on your permissions. You can ask questions, request summaries, check employee information, analyse absences, work with documents, or manage selected HR data — all through a natural language conversation.

What can you do with the Sloneek MCP server?

The Sloneek MCP server gives your AI assistant access to the same capabilities available through Sloneek Intelligence.

At the moment, the following agents are available:

• User agent
  Work with employee profiles, personal data, contracts, timeline events, and user-related information.
• Absence agent
  Check absence records, absence balances, requests, and absence-related summaries.
• Activity agent
  Analyse activities, time tracking, projects, clients, approval statuses, and related reports.
• Document agent
  Search documents, check ownership, review pending signatures or confirmations, and work with document-related information.
• Compensation agent
  Work with salary summaries, contract settings, bonuses, deductions, meal vouchers, working-hours records, and compensation-related data.

More agents will be added gradually, including agents for modules such as ATS and Performance.

Example: Managing contracts through ChatGPT

Once everything is connected, a typical conversation can look like this:

You: “Hey, whose contract is about to finish next month?”

ChatGPT: “Based on the information in Sloneek, these 3 users will finish their contracts next month: John Doe, Peter Poe, and Karin Jones. Would you like to extend any of them for another year?”

You: “Yeah, but only for Peter and Karin. John is leaving us.”

ChatGPT: “Sure, I’ve extended the contracts by one year for Peter Poe and Karin Jones. Anything else I can help you with?”

That is the main value of the Sloneek MCP server: you can work with HR data in a simple, conversational way.

Before you start

Before connecting Sloneek MCP to ChatGPT, or another assistant, make sure you have:

1. Access to Sloneek

   You need an active Sloneek account with the correct permissions.

2. Access to MCP tokens in Sloneek

   You need access to the Settings / MCP tokens section in Sloneek, where you can create a new MCP token for your AI tool.

3. A supported AI tool

   You need an MCP-compatible tool, such as ChatGPT or another client that supports MCP connections. 

   If you are using a company-managed AI workspace, such as a company ChatGPT or Gemini workspace, the Sloneek MCP server first needs to be added by the administrator of that AI workspace.

   The administrator connects and publishes the MCP server in the workspace settings. After that, individual users can find the published Sloneek MCP app in their AI tool, connect it to their account, and start using it.

4. The right permissions

   The assistant can only work with the data and actions your user account is allowed to access in Sloneek.

How to set up the Sloneek MCP server

The exact setup may vary depending on the AI tool you use. In company-managed AI workspaces, the setup usually has two parts:

1. The AI workspace administrator adds and publishes the Sloneek MCP server.
2. Individual users connect the published Sloneek MCP app in their own AI assistant.

Step 1: Admin adds Sloneek MCP to the AI workspace

First, the administrator of your company AI workspace needs to add Sloneek MCP as an app, connector, or MCP server.

For example, in ChatGPT, the admin can go to:

Workspace settings / Apps

Then click:

Create

The admin then creates a new app for Sloneek MCP.

Use a clear name such as:

Sloneek MCP

This name will later be visible to users when they search for the app in their AI assistant.

Step 2: Admin enters the Sloneek MCP server configuration

In the new app configuration, the admin enters the Sloneek MCP server details.

Use the following configuration:

MCP Server URL

https://api.server-agent.intelligence.sloneek.com/mcp

Authentication

Access token / API key

Header scheme

Custom Header

Header name

X-Api-Key

Step 3: Admin reviews and publishes the app

Before the app is published, the AI workspace may ask the admin to review and confirm potential risks and permissions.

For example, in ChatGPT, the admin may need to confirm that:

• they trust the app,
• they have reviewed the parameters being sent to each action,
• the app does not request sensitive data that it should not access,
• they understand the potential security implications.

After reviewing the configuration and permissions, the admin publishes the app to the company workspace.

Once published, the Sloneek MCP app becomes available to workspace users.

Step 4: Users connect Sloneek MCP in their AI assistant

After the admin publishes the Sloneek MCP app, individual users can connect it in their own AI assistant. 

In order to that, they need to obtain the MCP token from Sloneek. Please note: Only Sloneek admins can do this now. We plan to make Sloneek Intelligence available to regular users and managers later this year or next year.

In Sloneek, go to:

Settings / MCP tokens

Then click:

Create new token

Give your token a clear name, for example:

ChatGPT MCP
GPT Token

After creating the token, copy the token:

After that, depending on your AI tool, you need to insert that token to the right place.

For example, in ChatGPT, go to:

Apps

Then search for the app name (ask your workspace admin), for example:

Sloneek MCP

After selecting the app, connect it to your account by inserting the MCP token from Sloneek:

This token allows ChatGPT, or another MCP-compatible assistant to connect to Sloneek based on the permissions of the user account.

Only use this token in a trusted company-managed AI environment. Sloneek strongly recommends not connecting MCP tokens to personal AI assistants or personal accounts when working with sensitive employee data.

Step 5: Start using natural language

Once connected, you can start working with Sloneek directly through conversation.

For example:

“Show me all users with contracts ending in the next 30 days.”
“Which employees are currently absent?”
“Summarise pending document signatures.”
“Get me a salary analysis of our Sales team and compare it with the market standards.”
“Check compensation changes for this quarter.”

You do not need to know where every report is located or which filters to use. Just describe what you need.

Permissions and security

The Sloneek MCP server respects user permissions.

That means the assistant should only be able to access and perform actions that your Sloneek account is allowed to do. If you do not have access to certain data in Sloneek, your AI assistant should not have access to it either.

However, it is your responsibility to decide which AI tool you connect Sloneek to.

Sloneek strongly recommends connecting MCP only to trusted company-managed AI environments, such as your company ChatGPT, or another approved business AI workspace, especially when working with sensitive employee data.

This is especially important for HR data such as:

• personal employee information,

• compensation data,

• contracts,

• absence records,

• documents,

• performance-related information.

Before connecting Sloneek MCP to any AI tool, make sure that:

• the tool is approved for use in your company,

• the account is managed by your organisation,

• the data is processed according to your internal security policies,

• you understand what data may be sent to the AI assistant,

• and you only use the connection for authorised work purposes.

Always treat MCP tokens as sensitive credentials. Do not share them publicly, do not store them in unsecured places, and remove tokens that are no longer needed.

Read, update, and analyse Sloneek data

Depending on your permissions and the available agents, you can use the Sloneek MCP server for different types of tasks.

Reading data

Ask questions such as:

“Who is starting next month?”
“Which contracts end this quarter?”
“Who has pending documents to sign?”

Updating data

Ask the assistant to perform changes, for example:

“Extend Peter’s contract by one year.”
“Add a bonus for this employee: Karin Jones.”

For sensitive updates, confirmation may be required before the change is applied.

Analysing data

You can also ask for summaries and insights:

“Summarise absence trends for the last quarter.”
“Which teams logged the most activities last month?”
“Show me compensation changes from this year.”

Best practices

To get the best results, use clear and specific prompts.

Instead of:

“Show me contracts.”

Try:

“Show me all employees whose contracts end within the next 30 days.”

Instead of:

“Check absences.”

Try:

“Show me all approved absences for the Sales team next month.”

Instead of:

“Update Peter.”

Try:

“Extend Peter Poe’s contract end date by one year.”

The more specific your request is, the better and safer the result will be.

What is coming next?

The current version of the Sloneek MCP server supports the same core capabilities as Sloneek Intelligence.

At the moment, you can work with agents for users, absences, activities, documents, and compensation. More agents will be added gradually, including support for modules such as ATS and Performance.

This means the number of tasks you can complete directly from ChatGPT, and other MCP-compatible tools will continue to grow.

Summary

The Sloneek MCP server brings Sloneek directly into the tools you already use every day.

You can connect it to ChatGPT, or another MCP-compatible assistant and start working with HR data through natural language.

Ask questions, analyse data, update records, check contracts, review absences, inspect documents, and much more — without switching between multiple screens.

Sloneek MCP makes everyday HR work faster, simpler, and more conversational.