# AI assistants

> Connect Claude, ChatGPT, Cursor or another AI assistant to the Quenza help center, so its answers about Quenza come from these articles.

Source: https://quenza.com/help/ai-assistants/

AI assistants such as ChatGPT and Claude know a little about Quenza, but not always the current buttons, menus and plans. Connect them to the help center and they look up the answer in these articles before they reply, with a link to the page they used. It works on every plan and needs no login: the help center is public, and the assistant only reads it.

For example, Sarah Bennett asks Claude "How do I give a new client in my Mindfulness group the exercises the group already had?". Claude searches the help center, reads [New group members](https://quenza.com/help/clients/new-group-members/) and explains the setting, with the link.

The assistant only reads the help center. It cannot see or change anything in your Quenza workspace.

## Ask AI in the search

You don't need your own assistant to ask a question in your own words. Click **Ask AI** at the top of any page, or type your question in the search and press Enter. The answer is written from the help articles only, in the language you ask in, with numbered links to the articles it used. Nothing you ask is stored.

## Ask about one article

For a question about the article you are reading, open the **Copy page** menu at the top of the article:

- **Open in ChatGPT** or **Open in Claude** starts a chat that has read the article.
- **Copy page** copies the article as plain text, to paste into any assistant.
- **View as Markdown** opens the article as plain text.

## Connect the whole help center

The help center is also an MCP server. MCP (Model Context Protocol) is the standard way for AI assistants to use outside tools. Add this address as a custom connector or MCP server in your assistant:

```text
https://quenza.com/help/mcp
```

The assistant then gets three tools: search the help center, read an article, and list all articles. It needs no API key or login.

Where to add it depends on the assistant:

- **Claude** (claude.ai and the desktop app): add a custom connector with the address in your connector settings.
- **ChatGPT**: add a custom connector (also called an app) with the address. Depending on your ChatGPT plan, this needs developer mode.
- **Claude Code**: run

  ```bash
  claude mcp add --transport http quenza-help https://quenza.com/help/mcp
  ```

- **Cursor**: add the server to `.cursor/mcp.json` in your project, or to `~/.cursor/mcp.json` for all projects:

  ```json
  { "mcpServers": { "quenza-help": { "url": "https://quenza.com/help/mcp" } } }
  ```

- **VS Code**: add the server to `.vscode/mcp.json`:

  ```json
  { "servers": { "quenza-help": { "type": "http", "url": "https://quenza.com/help/mcp" } } }
  ```

The assistants change their settings pages from time to time; their own help explains where custom connectors are today.

> **Note:** An assistant can still make mistakes. When an answer matters, for example about billing or deleting data, open the linked article and check it. For anything the help center does not answer, email our support team at help@quenza.com.

## For developers

- [llms.txt](https://quenza.com/help/llms.txt) lists every article with its Markdown address; [llms-full.txt](https://quenza.com/help/llms-full.txt) contains all of them in one file.
- Every article is available as Markdown: add `.md` to its address without the last slash, such as `/help/clients/import-clients.md`.
- The MCP server speaks Streamable HTTP without sessions and offers the tools `search_quenza_help`, `get_quenza_help_page` and `list_quenza_help_pages`, all read-only.
- To work with your own workspace data, use the [Quenza API](https://quenza.com/help/api/) instead.
