Nurevo WebMCP

Description

Nurevo WebMCP Canary loads an external WebMCP tag on public WordPress pages. The tag reads public form structure in the browser and requests MCP, AEO, and autofill metadata from the configured Nurevo WebMCP server.

The plugin is designed for standard HTML forms and common WordPress form plugins.

Administrators can issue a site key, configure the hosted tag URL, and view a WebMCP dashboard. The dashboard shows collection progress while sample size is low, then measured form completion stats, benchmark panels, and real improvement suggestions once enough data exists.

The plugin also provides optional AEO (Answer Engine Optimization) features that help AI search engines and assistants understand your site: server-side schema.org structured data generated from your own content, an optional /llms.txt document, and optional AI-crawler allowances in robots.txt. All AEO features are controlled from WebMCP > Settings and are described under External Service Disclosure below.

External Service Disclosure

This plugin connects to an external service operated by Nurevo when enabled.

Default service endpoints:

• Service website: https://nurevo.jp/
• WebMCP / AEO Worker: https://nurevo.jp/
• Default script URL: https://nurevo.jp/tag.js

WebMCP tag data. Data sent to the WebMCP service:

• Public form structure, such as field selectors, field names, field types, labels, and option counts.
• Site host and page path.
• Site key used for authorization.
• Form execution outcome, such as success, failure, abandoned, or unknown.
• Validation metadata, such as error type, validity flags, sanitized error text, completed step, and duration.
• Browser-side WebMCP registration events needed to detect integration failures.

The WebMCP footprint flow does not send visitor-entered input values, names, email addresses, phone numbers, postal addresses, message bodies, payment details, or full page content.

AEO features (schema, llms.txt, robots.txt). When the optional AEO features are enabled, additional data is processed:

• Page metadata — the page title, meta description, Open Graph tags, canonical URL, language, headings (h1 to h3), and any existing JSON-LD — is sent to the Nurevo service so it can generate schema.org structured data for the page. Strings that look like email addresses or phone numbers are masked before this data leaves the page.
• The Nurevo service uses OpenAI to help generate this structured data from the submitted page metadata. See OpenAI’s privacy policy at https://openai.com/policies/privacy-policy/ and terms at https://openai.com/policies/ .
• When “Serve /llms.txt” is enabled, the plugin requests an llms.txt document from the Nurevo service and serves it at your site’s /llms.txt URL.
• When “Allow AI crawlers in robots.txt” is enabled, the plugin adds Allow rules for major AI crawlers (GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, anthropic-ai, PerplexityBot, Google-Extended, CCBot, Applebot-Extended) to WordPress’s virtual robots.txt.
• Server-side schema.org JSON-LD (Organization, WebSite, and the current page) is generated locally from your own WordPress content and printed in the page head. This step does not send data to any external service.

The service is intended to store privacy-safe structure, outcome, and page-metadata only. Site owners are responsible for confirming that their own use of external services complies with applicable privacy laws.

Service links:

• Privacy Policy: https://nurevo.jp/privacy
• Terms of Service: https://nurevo.jp/terms

Technical details