NIS2 Claude skill, connector or MCP server: which do you need?
A skill gives Claude instructions and knowledge for a task, such as NIS 2. A connector links Claude to another program; you add a custom one by the address of an MCP server. To read and fill in your NIS 2 walkthrough you connect its MCP server; your own skill can set out how Claude goes about it.
What is the difference between a skill, a connector and an MCP server?
According to Anthropic, a skill is a folder of instructions, scripts and resources that Claude loads when a task calls for it. At its core is a text file called SKILL.md with a name and a description of when Claude should use it.
An MCP server is a program's connection point for AI assistants. MCP (Model Context Protocol) is the open standard for it. The NIS 2 walkthrough has its MCP server at https://nisd2.eu/api/mcp.
According to Anthropic, a connector links Claude to an outside app or service, so Claude can find information there and take actions for you. You add a custom connector by the address of an MCP server. Anthropic sums up the difference like this: MCP connects Claude to external services and data sources, while skills provide the knowledge of how to complete a task.
What a NIS 2 skill can do, and what it cannot
A skill can explain to Claude what NIS 2 requires and set out how your company works: which lists it keeps, who is responsible for what, which terms it uses. It contains what someone wrote into it, and it goes out of date when the law or its transposition changes.
A skill brings no access to your data. Claude reads your progress in the walkthrough, your suppliers and your ratings through the MCP server.
Anthropic advises using skills only from trusted sources: ones you wrote yourself or got from Anthropic. A malicious skill can direct Claude to use tools or run code in ways that do not match its stated purpose. With the walkthrough connected, its tools are among them, so check a third-party NIS 2 skill file by file before you use it.
What the connector to the walkthrough does
Through the connector Claude reads your walkthrough: each item with its explanation, the provisions of the law it rests on and what your company has already entered. Then Claude fills in what the item asks for, as the person who connected it and with their rights.
Claude cannot finish, decide or sign anything there, because the server has no tool for it. That stays with you in the app, and the approval with management.
In Claude on the web and desktop you add a custom connector with the address and choose "Register automatically" for registration. In Claude Code you use the command below, then sign in with /mcp.
First an admin turns on "Allow AI assistants for this company" in Settings. We have tested Claude Code ourselves, Claude on the web not yet.
claude mcp add --transport http nisd2 https://nisd2.eu/api/mcpWhen do I need which?
You want to understand what NIS 2 requires, or have a text about the law drafted: a skill helps with that, without access to your data. Check the result against the text of the law.
You want to fill in your lists in the walkthrough, such as the inventory in 2.2: for that you connect the MCP server. Through it Claude enters data into your walkthrough, with a mark on every screen it wrote.
You can combine the two; that is our suggestion, and the walkthrough has no feature of its own for it. A skill your company writes itself sets out how Claude works with the connector, for example which list counts as the source for suppliers. The rule to leave open what your records do not show already comes with the server's instructions.
Do not put the key from Settings into a skill. It is your access for 90 days, in your name.
Where do I create a skill?
In Claude on the web, according to Anthropic, you upload a skill of your own as a ZIP file under "Customize", "Skills", via "+", "+ Create skill" and "Upload a skill". Code execution has to be on ("Code execution and file creation"). Check in your account whether your plan allows custom skills.
In Claude Code a skill is a folder with a SKILL.md, under ~/.claude/skills for yourself or under .claude/skills in the project. According to Anthropic, Claude Code picks up the skills from your Claude account when it is signed in with that account. Nothing syncs the other way or to the API; there you create a skill yourself.
How the walkthrough does it
The first items show where a skill stops and the connector starts.
2.1 "Decide how you rate risks": the item itself explains the method from BSI Standard 200-3, with a matrix and examples. Claude needs no skill for that; it reads the explanation through the connector.
On "Adopt the risk assessment method", Claude can adopt the method for your company through the connector, if that is what you want.
2.2 "Write down what your company has": with your inventory in the chat, Claude ticks on the four lists what applies to you and enters which software exactly it is and from whom. A skill of your own can set out which of your lists is the source.
Every screen Claude wrote shows the assistant's name and the day. You check the lists and finish the item in the app.
Common questions
Is there a NIS 2 skill from nisd2.eu?
No, not as of October 2026. nisd2.eu offers an MCP server for the walkthrough, which you connect in Claude as a connector. What NIS 2 requires for an item is written in the walkthrough beside that item, and Claude reads it through the connector.
Can a skill enter anything into the walkthrough?
Not without the connector, because a skill brings no access to the walkthrough. It can tell Claude to use the tools of the connected server. The entry then goes through the connector, with your rights and with the mark on the screen.
Is a connector the same as an MCP server?
Nearly. The MCP server runs at the program, here in the walkthrough at /api/mcp. The connector is the entry in Claude that connects to it; in Claude Code the same entry is simply called an MCP server.
What does Anthropic see when Claude reads the walkthrough?
What Claude reads through the connector goes to Anthropic. For commercial products such as Claude for Work, Anthropic says it acts as a processor and does not use inputs and outputs for training by default. Personal accounts come with their own terms.
Read next
- Claude custom connector: connecting the NIS2 walkthroughAs a custom connector with “Register automatically”. Which plan, who adds it, and why Claude needs an instance reachable from the internet.
- What is a NIS2 MCP server?The connection through which your own AI assistant reads and fills in the walkthrough, as you. What it does, what it may not do and how to start.
- Claude Code: adding the NIS2 walkthrough as an MCP serverOne command, one browser sign-in, and Claude Code reads and fills your walkthrough. With a key for scripts and a first task for 5.1.
Sources
- Anthropic help, What are skills? and Use skills in Claude, support.claude.com/en/articles/12512176 and 12512180, as of October 2026
- Anthropic, Agent Skills, platform.claude.com/docs/en/agents-and-tools/agent-skills/overview, and Claude Code skills, code.claude.com/docs/en/skills, as of October 2026
- Anthropic, Connectors, claude.com/docs/connectors/overview, and Add a connector that isn't in the directory, claude.com/docs/connectors/custom/remote-mcp, as of October 2026
- Anthropic Privacy Center, articles 7996868 (training) and 9267385 (processor role), privacy.claude.com
- open-isms, lib/mcp/server.ts and lib/mcp/tools.ts (what the server reads and fills in), github.com/NISD2/open-isms
In practice
Where this comes up in the walkthrough
The walkthrough turns this into steps you fill in, with the law explained beside each one. Open a step to see its screen, filled in for a made-up company.
Risk Management
Decide how you rate risks
We recommend a ready-made risk assessment method.
Risk Management
Write down what your company has
A list of the processes, software and devices you work with.
What the walkthrough costs
€4,800 net a year, plus VAT
30 days money backBy invoice, due in 30 days. Cancel within the first 30 days, before the first export or sign-off: invoice cancelled or money back.
Guided NIS 2 implementation, step by step
A history of changes and sign-offs to show your regulator
Deadlines and reminders for each requirement
Security questionnaire for your suppliers
Export to PDF or CSV at any time
Been asked to handle NIS2?
In the end you have everything in one document: what is done, what is still open, and your management's approval with name and date. The app writes the texts.
