OnPage.dev

Official20 tools

by seoonpage · JavaScript

Free SEO MCP that closes the loop: find, fix, ship as a PR or to WordPress, prove. 50 tools, no key.

dev.onpage/onpage — Model Context Protocol (MCP) Server

The dev.onpage/onpage MCP server is a free hosted SEO and AI-visibility scanner for agents. It provides capabilities to scan, fix, and verify SEO and visibility issues, including handling redirects and checking robots.txt. The server is intended for agent-driven workflows around website content and configuration.

🛠️ Key Features

Scan SEO/AI-visibility issues
Fix issues
Verify results
Redirects handling
robots.txt checks

🚀 Use Cases

Agent-assisted SEO audits
Automated remediation and re-checking
Validation of redirect behavior
Verification of robots.txt-related access rules

⚡ Developer Benefits

Designed for agent workflows (scan → fix → verify)
Hosted scanner for consistent visibility checks
Coverage includes redirects and robots.txt

⚠️ Limitations

Functionality described as scanning/fixing/verifying plus redirects and robots.txt; other behaviors are not specified.

Topics

ai-visibilityclaude-codellms-txtmcpseochatgptclaudemcp-serverseo-toolsaeogenerative-engine-optimizationgeoai-agentscore-web-vitalscursorllmmodel-context-protocolseo-audittechnical-seowordpress

Captured live from the server via tools/list.

scan_page

Scan one public web page for SEO and AI visibility. Returns a 0-100 score, the issues to fix first with why and how, AI readiness, and the key facts (title, description, headings, words, speed, structured data). Use this first for any question about how a page performs in Google or AI search.

Parameters1

url

string

required

Full address of the page, for example https://example.com/pricing. A bare domain like example.com also works.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address of the page, for example https://example.com/pricing. A bare domain like example.com also works."
    }
  },
  "required": [
    "url"
  ],
  "additionalProperties": false
}

check_ai_visibility

Check whether ChatGPT, Claude, Perplexity, Gemini and other AI crawlers may read a page (robots.txt per bot), whether llms.txt exists, how ready the content is for AI answers (11 checks), and how an AI model sees the page as plain text. Use this for questions like 'why does ChatGPT not mention my site'.

Parameters1

url

string

required

Full address of the page, for example https://example.com/pricing. A bare domain like example.com also works.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address of the page, for example https://example.com/pricing. A bare domain like example.com also works."
    }
  },
  "required": [
    "url"
  ],
  "additionalProperties": false
}

check_focus_keyword

Check how well a page targets one search term: title, meta description, H1, URL, first 100 words, subheadings, image alt texts and keyword density (flags stuffing over 3%). Returns 8 checks with a tip each.

Parameters2

url	string	required	Full address of the page, for example https://example.com/pricing. A bare domain like example.com also works.
keyword	string	required	The search term the page should rank for, for example 'running shoes'.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address of the page, for example https://example.com/pricing. A bare domain like example.com also works."
    },
    "keyword": {
      "type": "string",
      "description": "The search term the page should rank for, for example 'running shoes'."
    }
  },
  "required": [
    "url",
    "keyword"
  ],
  "additionalProperties": false
}

get_fix_pack

Generate ready-to-paste HTML for the issues on a page, built from the page's own content: a title tag, meta description, canonical, social (Open Graph) tags, viewport and a JSON-LD starting point. Only issues the page actually has are included.

Parameters1

url

string

required

Full address of the page, for example https://example.com/pricing. A bare domain like example.com also works.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address of the page, for example https://example.com/pricing. A bare domain like example.com also works."
    }
  },
  "required": [
    "url"
  ],
  "additionalProperties": false
}

scan_html

Scan raw HTML that is not published yet, for example a local build, a template or a draft from your editor. Returns the same score, fixes and AI readiness as scan_page. Pass the full HTML document; url is optional and only used to resolve relative links. Use this in coding assistants to check SEO before deploying.

Parameters2

html	string	required	The complete HTML document, up to 900 KB.
url	string	optional	Optional address the page will live at, used for relative links and the canonical check.

Raw schema

{
  "type": "object",
  "properties": {
    "html": {
      "type": "string",
      "description": "The complete HTML document, up to 900 KB."
    },
    "url": {
      "type": "string",
      "description": "Optional address the page will live at, used for relative links and the canonical check."
    }
  },
  "required": [
    "html"
  ],
  "additionalProperties": false
}

rescan_and_compare

Scan a live page again and compare it with the previous scan of that page: score change, issues fixed, new issues and AI readiness change. Use this after deploying fixes to confirm they worked, or to catch regressions such as a robots.txt change that suddenly blocks AI crawlers. Repeat until the score is where you want it.

Parameters1

url

string

required

Full address of the page, for example https://example.com/pricing. A bare domain also works.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address of the page, for example https://example.com/pricing. A bare domain also works."
    }
  },
  "required": [
    "url"
  ],
  "additionalProperties": false
}

check_snippet

Estimate how a title and meta description show in Google: pixel width against the cut-off on desktop and mobile, and the truncated text a searcher would see. No network call. Use it to iterate on title and description variants until they fit.

Parameters2

title	string	required	Proposed title tag text.
description	string	optional	Proposed meta description.

Raw schema

{
  "type": "object",
  "properties": {
    "title": {
      "type": "string",
      "description": "Proposed title tag text."
    },
    "description": {
      "type": "string",
      "description": "Proposed meta description."
    }
  },
  "required": [
    "title"
  ],
  "additionalProperties": false
}

generate_schema

Generate a JSON-LD block for a page from its own content and validate it against Google's rich result rules. Types: Article, Product, FAQPage (built from question headings and their answers), Organization (with sameAs links found on the page), BreadcrumbList (from the URL), or auto to pick the best fit. Values that cannot be read from the page are marked TODO.

Parameters2

url	string	required	Full address of the page, for example https://example.com/pricing. A bare domain also works.
type	string	optional	Schema type, default auto.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address of the page, for example https://example.com/pricing. A bare domain also works."
    },
    "type": {
      "type": "string",
      "enum": [
        "auto",
        "Article",
        "Product",
        "FAQPage",
        "Organization",
        "BreadcrumbList"
      ],
      "description": "Schema type, default auto."
    }
  },
  "required": [
    "url"
  ],
  "additionalProperties": false
}

find_answer_passages

Find the passages on a page that best answer a given question, the way AI answer engines pick text to quote. Returns the top passages with their heading, word count and whether they are a quotable length (40 to 90 words). Use the result to judge whether the page answers the question well, and to rewrite the best passage into a direct answer.

Parameters2

url	string	required	Full address of the page, for example https://example.com/pricing. A bare domain also works.
question	string	required	The question people ask, for example 'how long does shipping take to Germany'.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address of the page, for example https://example.com/pricing. A bare domain also works."
    },
    "question": {
      "type": "string",
      "description": "The question people ask, for example 'how long does shipping take to Germany'."
    }
  },
  "required": [
    "url",
    "question"
  ],
  "additionalProperties": false
}

generate_llms_txt

Build a ready-to-upload llms.txt for a site from its sitemap and home page: site name, summary and the main pages grouped by section. Upload the result to the root of the site, for example https://example.com/llms.txt.

Parameters1

url

string

required

Home page of the site, for example https://example.com.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Home page of the site, for example https://example.com."
    }
  },
  "required": [
    "url"
  ],
  "additionalProperties": false
}

ai_crawler_policy

Write robots.txt rules for AI crawlers based on a policy, and merge them into the site's current robots.txt. Policies: allow_all, search_only (allow AI search and user-requested fetches, block model training), block_training (same as search_only), block_all. Returns the full new robots.txt and what changes.

Parameters2

url	string	required	Any page of the site.
policy	string	required	Which AI crawlers to allow.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Any page of the site."
    },
    "policy": {
      "type": "string",
      "enum": [
        "allow_all",
        "search_only",
        "block_training",
        "block_all"
      ],
      "description": "Which AI crawlers to allow."
    }
  },
  "required": [
    "url",
    "policy"
  ],
  "additionalProperties": false
}

deep_audit

Everything OnPage.dev measures on a page, by section: speed hints from the HTML, links and anchor texts, accessibility basics, image SEO, rich results, security headers, content (words, headings, readability, keywords) and technical facts (status, canonical, robots, redirects, hreflang). Pick sections to keep the answer short, or leave them out for all. Every issue has a stable code you can track across scans.

Parameters2

url	string	required	Full address of the page, for example https://example.com/pricing. A bare domain also works.
sections	array	optional	Sections to include. Default: all.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address of the page, for example https://example.com/pricing. A bare domain also works."
    },
    "sections": {
      "type": "array",
      "items": {
        "type": "string",
        "enum": [
          "speed",
          "links",
          "accessibility",
          "images",
          "rich_results",
          "security",
          "content",
          "technical"
        ]
      },
      "description": "Sections to include. Default: all."
    }
  },
  "required": [
    "url"
  ],
  "additionalProperties": false
}

check_urls

Check up to 20 URLs at once: final status code, the full redirect chain and where each one ends up. Built for site migrations, broken link sweeps and launch checks, for example 'do all old URLs 301 to the right new page?'. Pass expected targets to verify each redirect.

Parameters2

urls	array	required	Addresses to check.
expected	array	optional	Optional expected final address for each URL, in the same order.

Raw schema

{
  "type": "object",
  "properties": {
    "urls": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "minItems": 1,
      "maxItems": 20,
      "description": "Addresses to check."
    },
    "expected": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "maxItems": 20,
      "description": "Optional expected final address for each URL, in the same order."
    }
  },
  "required": [
    "urls"
  ],
  "additionalProperties": false
}

test_robots

Test a URL against the site's live robots.txt for one or more crawlers, using Google's matching rules (most specific user-agent group, longest matching rule, Allow wins ties, * and $ wildcards). Returns allowed or blocked per crawler and the exact rule that decides it. Optionally test a proposed robots.txt instead of the live one.

Parameters3

url	string	required	Full address of the page, for example https://example.com/pricing. A bare domain also works.
user_agents	array	optional	Crawler names, for example Googlebot, GPTBot, ClaudeBot, PerplexityBot. Default: Googlebot, Bingbot and the main AI crawlers.
robots_txt	string	optional	Optional robots.txt content to test instead of the live file.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address of the page, for example https://example.com/pricing. A bare domain also works."
    },
    "user_agents": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "maxItems": 20,
      "description": "Crawler names, for example Googlebot, GPTBot, ClaudeBot, PerplexityBot. Default: Googlebot, Bingbot and the main AI crawlers."
    },
    "robots_txt": {
      "type": "string",
      "description": "Optional robots.txt content to test instead of the live file."
    }
  },
  "required": [
    "url"
  ],
  "additionalProperties": false
}

validate_schema

Validate JSON-LD you wrote before you publish it: JSON syntax, @context and @type, absolute URLs, ISO dates, and Google's required and recommended fields per rich result type. Accepts a JSON-LD object, an array, a @graph, or a full <script type="application/ld+json"> block.

Parameters1

jsonld

string

required

The JSON-LD as text, with or without the <script> tag.

Raw schema

{
  "type": "object",
  "properties": {
    "jsonld": {
      "type": "string",
      "description": "The JSON-LD as text, with or without the <script> tag."
    }
  },
  "required": [
    "jsonld"
  ],
  "additionalProperties": false
}

compare_html

Compare two versions of a page's HTML, for example before and after a code change, and report what got better or worse for SEO: score, issues fixed and introduced (with stable codes), and changes to title, description, H1, canonical, robots, structured data and word count. Use it in code review to catch SEO regressions before they ship.

Parameters3

before	string	required	HTML before the change.
after	string	required	HTML after the change.
url	string	optional	Optional address the page lives at.

Raw schema

{
  "type": "object",
  "properties": {
    "before": {
      "type": "string",
      "description": "HTML before the change."
    },
    "after": {
      "type": "string",
      "description": "HTML after the change."
    },
    "url": {
      "type": "string",
      "description": "Optional address the page lives at."
    }
  },
  "required": [
    "before",
    "after"
  ],
  "additionalProperties": false
}

start_site_audit

Start an SEO audit of a whole site (up to 25 pages from the sitemap, or from the home page links when there is no sitemap). Returns an audit_id and the first progress. Then call get_site_audit with that id until status is complete; each call scans the next few pages. The finished audit gives the issues across the site by code with the affected pages, broken pages, duplicate titles and descriptions, and orphan pages, and it powers suggest_internal_links.

Parameters2

url	string	required	Full address, for example https://example.com. A bare domain also works.
max_pages	integer	optional	How many pages to audit, default 10, at most 25.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address, for example https://example.com. A bare domain also works."
    },
    "max_pages": {
      "type": "integer",
      "minimum": 2,
      "maximum": 25,
      "description": "How many pages to audit, default 10, at most 25."
    }
  },
  "required": [
    "url"
  ],
  "additionalProperties": false
}

get_site_audit

Continue a site audit started with start_site_audit: each call scans the next pages and reports progress. Keep calling until status is complete, then you get the site-wide results. Audits are kept for one hour.

Parameters1

audit_id

string

required

The id returned by start_site_audit.

Raw schema

{
  "type": "object",
  "properties": {
    "audit_id": {
      "type": "string",
      "description": "The id returned by start_site_audit."
    }
  },
  "required": [
    "audit_id"
  ],
  "additionalProperties": false
}

suggest_internal_links

Suggest which existing pages should link to a target page, with anchor text, based on a completed site audit. The target can be a page from the audit or a new page you are writing (give its topic). Pages that already link to the target are left out, and orphan pages are flagged. Use it when publishing new content or fixing pages that get no internal links.

Parameters3

audit_id	string	required	Id of a completed site audit.
target_url	string	optional	Page that should receive links. Can be a new URL that is not live yet.
topic	string	optional	Topic or main keyword of the target page. Needed for a new page; optional for an audited page.

Raw schema

{
  "type": "object",
  "properties": {
    "audit_id": {
      "type": "string",
      "description": "Id of a completed site audit."
    },
    "target_url": {
      "type": "string",
      "description": "Page that should receive links. Can be a new URL that is not live yet."
    },
    "topic": {
      "type": "string",
      "description": "Topic or main keyword of the target page. Needed for a new page; optional for an audited page."
    }
  },
  "required": [
    "audit_id"
  ],
  "additionalProperties": false
}

compare_pages

Compare a page with up to 3 competitor pages that rank for the same search: score, AI readiness, words, headings, rich results, speed and more side by side, plus per competitor the issues to fix to catch up, structured data they have, words they use that the page does not (content gaps), and where the page is ahead. It scans one page per call and keeps results for 15 minutes, so with several new pages it asks you to call it again with the same arguments until the comparison is complete.

Parameters3

url	string	required	Full address, for example https://example.com. A bare domain also works.
competitors	array	required	Competitor page addresses.
keyword	string	optional	Optional search term to compare keyword use.

Raw schema

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "Full address, for example https://example.com. A bare domain also works."
    },
    "competitors": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "minItems": 1,
      "maxItems": 3,
      "description": "Competitor page addresses."
    },
    "keyword": {
      "type": "string",
      "description": "Optional search term to compare keyword use."
    }
  },
  "required": [
    "url",
    "competitors"
  ],
  "additionalProperties": false
}

README.md15 KBView on GitHub

OnPage.dev MCP server

The SEO scanner your AI assistant uses to fix your site itself.
Scan, fix, check before deploy, confirm after. For Google and for AI search.

Free, no API key 20 tools

Install · Built for agents · Tools · Why it is different · Claude Code plugin · GitHub Action · onpage.dev

Most SEO tools give you a report. OnPage.dev gives your assistant something to do: write the fix, check it before you publish, and confirm it worked. It also checks what AI search sees, not only Google: which AI crawlers may read your site, whether your content is ready to be quoted, and whether you have an llms.txt.

code

You:        Fix the SEO on our pricing page and keep going until it scores 90 or more.

Assistant:  scan_page            46/100  no meta description, no H1, title 14 characters
            get_fix_pack         title, description and canonical, written from the page
            generate_schema      Product JSON-LD, 2 values to fill in
            (edits pricing.html)
            scan_html            92/100  before deploying, all errors gone
            rescan_and_compare   94/100  live, +48 since the first scan

            Done. 46 to 94. Two prices in the JSON-LD still need your real values.

Built for agents

What makes OnPage.dev different from every other SEO MCP server:

Stable issue codes and typed results. Every issue has a fixed code like meta-missing or h1-missing, and the main tools declare an output schema. An agent can work through issues one by one and prove each one is gone.
It checks the code the agent writes. scan_html, validate_schema and compare_html test new HTML and JSON-LD before it ships, and catch regressions such as a stray noindex or a removed canonical.
Launch and migration built in. check_urls follows every redirect chain for up to 20 URLs at once; test_robots answers "may this bot fetch this URL" with Google's own matching rules, including * and $ wildcards.
Whole sites, step by step. start_site_audit runs as a job the agent works through, so a 25-page audit fits a free hosted service. It reports issues across the site by code, broken and orphan pages and duplicates.
You vs the pages that rank. compare_pages puts a page next to up to 3 competitors with content gaps and fixes to catch up; suggest_internal_links shows which pages should link where.
No made-up advice. The onpage://rules resource gives the assistant every check, threshold and fix, so it explains SEO from the same rules the scanner uses.
Free and hosted. Paste one URL. No account, no API key, no install.

Install in one minute

The server is hosted. There is nothing to install or run, and no key.

Server URL: https://onpage.dev/mcp

Claude Code

bash

claude mcp add --transport http onpage https://onpage.dev/mcp

Or install the plugin for the server plus ready commands and skills.

Claude (claude.ai and desktop)

Settings → Connectors → Add custom connector → paste https://onpage.dev/mcp → save. Turn the connector on in a chat.

ChatGPT

Add a custom connector with https://onpage.dev/mcp (Settings → Connectors, with developer mode on). Which plans can use custom connectors is up to OpenAI.

Cursor

~/.cursor/mcp.json or .cursor/mcp.json in your project:

json

{
  "mcpServers": {
    "onpage": { "url": "https://onpage.dev/mcp" }
  }
}

VS Code

.vscode/mcp.json:

json

{
  "servers": {
    "onpage": { "type": "http", "url": "https://onpage.dev/mcp" }
  }
}

Any other MCP client

Streamable HTTP transport at https://onpage.dev/mcp. JSON responses, no authentication.

What it can do

20 tools in five jobs. Every result links to the full visual report on onpage.dev. Issues carry stable codes; the full list is in the onpage://rules resource.

Audit and fix

Tool	What it does	Example result
`scan_page`	Score from 0 to 100, the issues to fix first with why and how, AI readiness and key facts	`46/100, 5 fixes`
`deep_audit`	Everything measured, by section: speed hints, links and anchor texts, accessibility, image SEO, rich results, security headers, content and technical facts	`8 sections`
`get_fix_pack`	Ready-to-paste HTML for every issue, written from the page's own content	`<title>`, `<meta>`, canonical, social tags
`generate_schema`	Article, Product, FAQ, Organization or breadcrumb JSON-LD, validated against Google's rich result rules. Values it cannot read are marked TODO, never invented	`Product: eligible`
`rescan_and_compare`	Scan again after a fix: score change, what got fixed, what is new	`46 → 94`

Whole site and competitors

Tool	What it does	Example result
`start_site_audit`	Audit up to 25 pages from the sitemap as a job. Returns an `audit_id` and the first progress	`3 of 25 scanned`
`get_site_audit`	Continue until complete, then read the site-wide results: issues by code with affected pages, broken, orphan and duplicate pages	`25 pages, average 92`
`suggest_internal_links`	Which audited pages should link to a page, with anchor text. Works for a new page by topic	`3 links to add`
`compare_pages`	A page next to up to 3 competitors: side by side, content gaps, structured data they have and fixes to catch up. Flags cookie walls	`You lead on 12 of 14`

AI search

Tool	What it does	Example result
`check_ai_visibility`	Which AI crawlers may read the page (GPTBot, ClaudeBot, PerplexityBot, Google-Extended and more), llms.txt, 11 checks for AI answers, and the page as a model reads it	`7 of 8 crawlers allowed`
`find_answer_passages`	Does the page answer a question well enough for AI to quote it? Returns the best passages with length and fit	`Best passage, 64 words`
`ai_crawler_policy`	robots.txt rules for 13 AI crawlers from a policy (allow all, AI search only, block all), merged into your existing file	`13 bots, merged`
`generate_llms_txt`	A ready `llms.txt` built from the sitemap and home page	`llms.txt, 42 pages`

Launch and migration

Tool	What it does	Example result
`check_urls`	Status codes and full redirect chains for up to 20 URLs, checked against where each should land. Flags chains and temporary redirects	`19 of 20 OK`
`test_robots`	May Googlebot, GPTBot, ClaudeBot or any crawler fetch this URL? Uses Google's matching rules and returns the exact deciding rule. Can test a proposed robots.txt too	`Blocked by "Disallow: /search?"`

Writing and code

Tool	What it does	Example result
`scan_html`	Scan HTML that is not live yet: a local build, a template or a draft. Same score and fixes as `scan_page`	`92/100 before deploy`
`compare_html`	SEO diff between two versions of a page: score, issues fixed and introduced (by code), and changes to title, description, H1, canonical, robots and structured data	`72 → 18, noindex added`
`validate_schema`	Check JSON-LD you wrote: syntax, @context and @type, ISO dates, absolute URLs, placeholders and Google's required fields	`2 errors, 3 warnings`
`check_focus_keyword`	Eight checks for one search term: title, description, H1, URL, opening text, subheadings, alt texts and keyword use	`6 of 8`
`check_snippet`	Pixel-width estimate of title and description against Google's cut-off, desktop and mobile	`612 px, cut off`

Ready workflows (MCP prompts)

One click in clients that show prompts:

Full SEO and AI audit: scan, AI check and fix pack, then a prioritised plan.
Why can't AI find my site?: which assistants can read and quote you, and what to change.
Launch checklist: a tick or cross for everything a page needs before going live.
AI visibility makeover: robots.txt, llms.txt and answer passages in one go.
Pre-deploy SEO gate: scan the built HTML and block the release below a minimum score.

Reference resource

onpage://rules: every check with its stable code, why it matters and how to fix it, plus the key thresholds (title and description length in pixels, word counts, alt text, server response, AI answer length).

Score card in the chat

In clients that support MCP Apps (Claude) or the Apps SDK (ChatGPT), scan results also show as a visual card:

OnPage.dev score card in the chat: score 46, AI readiness 27, and the fixes to make first

Why it is different

	OnPage.dev	Data vendor MCPs (Ahrefs, Semrush and similar)	Self-hosted audit servers
Price	Free	Paid plan or API credits	Free
Setup	Paste one URL	Account and key	Install and run yourself, often extra API keys
Built for	Acting on results: fix, check before deploy, confirm	Research data: keywords, backlinks, rankings	Audits and reports
Scan unpublished HTML	Yes, `scan_html` and `compare_html`	No	Rarely
Agent-grade output	Stable issue codes, output schemas, rules resource	Data rows	Varies
Migration checks	Redirect chains, robots.txt tester	Partly	Partly
Whole-site audit	Up to 25 pages, free, as a job	Yes, paid	Yes, self-run
Competitor gaps	Page vs page, with content gaps	Keyword and backlink gaps	Rarely
AI search tools	Crawler access, AI-answer checks, robots.txt policy, llms.txt	AI visibility tracking on some	Rarely
Ready code	Fix pack, JSON-LD, robots.txt, llms.txt	No	Varies

What OnPage.dev does not do: search volumes, rankings or backlinks. For those, the data vendors are still the place to go, and they work well alongside OnPage.dev.

Example prompts

code

Why does ChatGPT never mention example.com? Check it with OnPage.dev.
Scan our pricing page and give me the five fixes with the code.
Let ChatGPT search read our site but block AI training. Write the robots.txt.
We moved to a new URL structure. Check that these 20 old URLs redirect to the right new pages.
Review this pull request for SEO regressions: compare the old and new HTML of the changed pages.
Audit our whole site and tell me the three fixes that help the most pages at once.
We are writing a page about trail running shoes. Which of our pages should link to it?
How does our pricing page compare with these two competitors, and what are we missing?
Does our blog post answer "how long does shipping to Germany take" well enough for AI to quote?
Write three title options for this page and check which ones fit in Google.
Fix the SEO issues in this project, check the build with OnPage.dev and keep going until it scores 95.

Claude Code plugin

The server plus commands and skills that run the fix loop for you.

code

/plugin marketplace add seoonpage/onpage-tools
/plugin install onpage@onpage-dev

/seo-check [url or path]: scan a live page, or find this project's built HTML and scan that, then fix what is wrong.
/ai-visibility: check whether ChatGPT, Claude, Perplexity and Gemini can read and quote the site, and write robots.txt and llms.txt.
seo-fix-loop skill: fixes in your source (templates, Next.js metadata, Astro frontmatter), checks with scan_html before deploy and confirms with rescan_and_compare after.

About 300 tokens of always-on context.

GitHub Action: SEO gate

Scans the HTML your build produces on every pull request, comments the scores and top fixes, and fails when a page drops below your minimum.

yaml

name: SEO gate
on: pull_request
permissions:
  contents: read
  pull-requests: write
jobs:
  seo:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v5
      - run: npm ci && npm run build
      - uses: seoonpage/onpage-tools@v1
        with:
          paths: "dist/**/*.html"          # built HTML to check
          base-url: "https://example.com"   # optional, for canonical and relative links
          min-score: "80"                   # fail below this
          max-files: "10"                   # about 20 scans per minute

The pull request gets a comment like this, updated on every push:

❌ OnPage.dev SEO gate: lowest score 36 (minimum 80)

Page Score AI readiness Fix first
dist/about/index.html 36 0 There is no meta description; There is no H1; The title is 5 characters
dist/index.html 90 22 Thin content: 32 words; No structured data; Open Graph is incomplete

Security and privacy

Read-only. Every tool only reads public pages or the HTML you send. Nothing is changed anywhere.
Public pages only. Private, internal and local network addresses are refused (SSRF protection), and so are non-web schemes.
No storage of content. HTML sent to scan_html is analysed in memory and dropped. For change tracking, only the score and issue names are kept, under a hashed key, for up to 30 days.
Prompt injection aware. Text from scanned pages is returned as data and marked as such.
Fair use limits. About 20 scans per minute per connection, plus a shared cap. No accounts, no keys, no tracking cookies.

Full details: onpage.dev/privacy and onpage.dev/terms.

Limits

Reads the HTML a server returns. JavaScript is not run, so content that only appears in the browser is not seen.
Link checks, image sizes and AI crawler access need a live URL, so scan_html skips them.
Pixel widths in check_snippet are an estimate; Google can also rewrite snippets.

License

MIT

Install

Tools (20)

scan_pagecheck_ai_visibilitycheck_focus_keywordget_fix_packscan_htmlrescan_and_comparecheck_snippetgenerate_schema

Remote endpoint

Streamable HTTP

Hosted server - connect over the network, no local install.

url

https://onpage.dev/mcp

Add to VS Code Add to Cursor

claude_desktop_config.json

json

{
  "mcpServers": {
    "onpage": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://onpage.dev/mcp"
      ]
    }
  }
}

Desktop config is stdio-only; this bridges via mcp-remote. Native remote: Settings > Connectors.

Page	Score	AI readiness	Fix first
`dist/about/index.html`	36	0	There is no meta description; There is no H1; The title is 5 characters
`dist/index.html`	90	22	Thin content: 32 words; No structured data; Open Graph is incomplete

dev.onpage/onpage — Model Context Protocol (MCP) Server

🛠️ Key Features

🚀 Use Cases

⚡ Developer Benefits

⚠️ Limitations

Topics

scan_page

check_ai_visibility

check_focus_keyword

get_fix_pack

scan_html

rescan_and_compare

check_snippet

generate_schema

find_answer_passages

generate_llms_txt

ai_crawler_policy

deep_audit

check_urls

test_robots

validate_schema

compare_html

start_site_audit

get_site_audit

suggest_internal_links

compare_pages

OnPage.dev MCP server

Built for agents

Install in one minute

What it can do

Audit and fix

Whole site and competitors

AI search

Launch and migration

Writing and code

Ready workflows (MCP prompts)

Reference resource

Score card in the chat

Why it is different

Example prompts

Claude Code plugin

GitHub Action: SEO gate

❌ OnPage.dev SEO gate: lowest score 36 (minimum 80)

Security and privacy

Limits

Links

License

Install

Remote endpoint

Related servers