> ## Documentation Index
> Fetch the complete documentation index at: https://help.weavely.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Weavely.ai MCP Server

> Use our MCP server to connect Weavely to your favourite AI and LLM tools!

<span className="weavely-jsonld" hidden>
  {`{
    "@context": "https://schema.org",
    "@graph": [
      {
        "@type": "WebPage",
        "@id": "https://help.weavely.ai/integrations/mcp",
        "url": "https://help.weavely.ai/integrations/mcp",
        "name": "Weavely.ai MCP Server",
        "description": "Use our MCP server to connect Weavely to your favourite AI and LLM tools!",
        "inLanguage": "en",
        "isPartOf": {
          "@type": "WebSite",
          "@id": "https://help.weavely.ai/#website",
          "name": "Weavely Help Center"
        },
        "breadcrumb": {
          "@type": "BreadcrumbList",
          "itemListElement": [
            { "@type": "ListItem", "position": 1, "name": "Integrations", "item": "https://help.weavely.ai/integrations" },
            { "@type": "ListItem", "position": 2, "name": "Weavely MCP", "item": "https://help.weavely.ai/integrations/mcp" }
          ]
        },
        "publisher": { "@id": "https://www.weavely.ai/#org" }
      },
      {
        "@type": "TechArticle",
        "headline": "Weavely.ai MCP Server",
        "description": "Weavely's MCP Server allows AI clients like Claude, Cursor, or others to build and publish forms conversationally—using the Model Context Protocol (MCP).",
        "inLanguage": "en",
        "author": { "@id": "https://www.weavely.ai/#org" },
        "publisher": { "@id": "https://www.weavely.ai/#org" },
        "mainEntityOfPage": { "@id": "https://help.weavely.ai/integrations/mcp" }
      },
      {
        "@type": "Organization",
        "@id": "https://www.weavely.ai/#org",
        "name": "WEAVE.LY BV",
        "url": "https://www.weavely.ai",
        "address": {
          "@type": "PostalAddress",
          "streetAddress": "Herbert Hooverlaan 141",
          "addressLocality": "Schaarbeek",
          "postalCode": "1030",
          "addressCountry": "BE"
        },
        "identifier": "BE0773.878.569",
        "sameAs": [
          "https://www.instagram.com/weavely.ai/",
          "https://www.linkedin.com/company/weave-ly/",
          "https://www.youtube.com/@weavely"
        ]
      }
    ]
    }`}
</span>

<iframe src="https://www.youtube.com/embed/NtS2UgeADhQ" title="YouTube video player" frameborder="0" className="w-full aspect-video rounded-xl" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen />

Weavely's MCP Server allows AI clients like **Claude**, **Cursor**, **Windsurf**, or any MCP-compatible agent to **build, style, and publish forms conversationally** — using standard [Model Context Protocol (MCP)](https://modelcontextprotocol.io/) tools.

This guide walks you through what the Weavely MCP is, how it works, and how to configure it for any compatible AI client.

## What is an MCP Server?

MCP (Model Context Protocol) is a standard that allows AI models to **call tools** — like APIs or custom functions — by passing structured arguments.

Instead of hardcoding every integration, you provide a reference to an MCP server that **exposes tools dynamically**. The LLM then learns what tools are available and when to use them, all without manual plumbing.

## What Can You Do with Weavely MCP?

The Weavely MCP server gives your AI client **full control over the form-building process** — from creating an empty form to adding questions, styling it, adding conditional logic, and publishing.

The AI builds the form step by step while you follow along via a **live preview URL** that updates in real-time. When you're happy with the result, ask the AI to publish and you'll get a link to claim ownership.

For example:

> "Create a post-purchase NPS survey for new customers. Use a 0–10 rating scale, add a follow-up text area that only shows when the score is below 7, and style it with a clean dark theme."

The AI will:

1. Create the form and share a live preview link
2. Add a scale rating element and a conditional text area
3. Set up logic to show the text area only for low scores
4. Apply a dark color theme with clean layout
5. Publish when you ask — giving you an editor URL to claim full ownership

## Setup

Add the Weavely MCP server to your AI client's configuration:

```json theme={null}
{
  "mcpServers": {
    "weavely": {
      "url": "https://mcp.weavely.ai/mcp"
    }
  }
}
```

This works with any client that supports **Streamable HTTP** MCP servers, including Claude Desktop, Claude Code, and Cursor.

<Note>
  If your client only supports stdio-based MCP servers, use the `mcp-remote` bridge:

  ```json theme={null}
  {
    "mcpServers": {
      "weavely": {
        "command": "npx",
        "args": [
          "mcp-remote",
          "https://mcp.weavely.ai/mcp"
        ]
      }
    }
  }
  ```

  `npx` must be available in your terminal. On Windows with a local agent, you may need to prefix with `cmd /c`.
</Note>

## Tools

The MCP server exposes **13 tools** that cover the full form-building lifecycle:

### Form management

| Tool               | Description                                                                           |
| ------------------ | ------------------------------------------------------------------------------------- |
| `create_form`      | Create a new form. Returns a `formId` and live preview URL. **Must be called first.** |
| `get_form_summary` | Get the current form structure — pages, elements, theme, settings                     |
| `publish_form`     | Publish the form and get an editor URL to claim ownership                             |

### Elements

| Tool               | Description                                           |
| ------------------ | ----------------------------------------------------- |
| `add_element`      | Add a question or content element to a page           |
| `update_element`   | Update an element's label, description, options, etc. |
| `remove_element`   | Remove an element                                     |
| `reorder_elements` | Move an element to a different position or page       |

### Pages

| Tool            | Description                           |
| --------------- | ------------------------------------- |
| `add_page`      | Add a new page (for multi-step forms) |
| `remove_page`   | Remove a page and its elements        |
| `reorder_pages` | Reorder form pages                    |

### Styling & configuration

| Tool           | Description                                        |
| -------------- | -------------------------------------------------- |
| `set_theme`    | Set colors, fonts, layout, and component styles    |
| `set_settings` | Configure language, progress bar, auto-save, etc.  |
| `set_logic`    | Add conditional show/hide rules and event triggers |

### Supported element types

The `add_element` tool supports **25 element types**:

* **Input fields**: text, number, email, phone, URL, time, date, text area
* **Choice fields**: radio buttons, checkboxes, dropdown, ranking
* **Image choice**: multiple-choice with images (images assigned on platform after publishing)
* **Matrix**: grid questions with rows and columns
* **Rating**: star rating, scale rating, range slider
* **Display**: headings, paragraphs (with HTML support)
* **Media**: embedded HTML, audio, video
* **Special**: file upload, signature, single checkbox

## How It Works

Once the MCP server is connected, the AI sees all 13 tools and uses them automatically based on your requests.

**Example conversation:**

> **You:** Create a customer feedback form with a star rating, a comment box, and an email field. Make it look professional.
>
> **AI:** I'll create that for you. Here's your live preview: `https://forms.weavely.ai/abc-123` — you can open it now and refresh to see updates as I build.
>
> *(AI calls create\_form, then add\_element three times, then set\_theme)*
>
> **You:** Add a dropdown asking how they found us, with options for Google, Social Media, and Friend.
>
> *(AI calls add\_element with radio-buttons and options)*
>
> **You:** Looks great, publish it!
>
> *(AI calls publish\_form)*
>
> **AI:** Your form is published! Open the editor to claim it: `https://forms.weavely.ai/editor/xyz-456?publish=true`

The live preview updates with every change — you just refresh the page to see the latest version.

## After Publishing

Once you publish and claim your form on the Weavely platform, you get access to additional features that aren't available through the MCP:

* **Integrations**: Make, Zapier, n8n, Google Sheets, HubSpot, Airtable, Notion
* **Email notifications**: get notified on submissions or send confirmation emails
* **Social media preview**: customize Open Graph image, title, and description
* **Custom domain**: serve the form on your own domain
* **Embed codes**: embed the form on any website
* **Analytics**: view submissions and response data

For the full developer reference, see our [Forms API documentation](https://help.weavely.ai/developers/forms) and [llms.txt](https://help.weavely.ai/llms.txt).