On-Page SEO Auditor - Titles, Meta, Headings, Schema & Issues

Audit the on-page SEO of any list of URLs. The actor fetches each page and reports titles, meta descriptions, canonicals, headings, Open Graph tags, structured data, link counts, image alt coverage and a plain-language list of the issues it found.

No API key, no proxy needed.

What you get per page

Response - statusCode, finalUrl, redirected, loadTimeMs, pageSizeBytes

Indexing and meta - title with titleLength, metaDescription with metaDescriptionLength, canonicalUrl, robotsMeta, isIndexable, lang, charset, hasViewport

Structure - h1 and h2 contents, h1Count, headingCounts for H1 to H6, wordCount

Links and media - internalLinkCount, externalLinkCount, externalDomains, imageCount, imagesMissingAlt

Social and structured data - ogTitle, ogDescription, ogImage, ogType, twitterCard, schemaTypes from JSON-LD, hreflangs

Findings - issues, a ready-to-read list such as "Title longer than 60 characters", "No H1 heading", "3 image(s) without alt text"

Input

• urls - the pages to audit, separated by commas, spaces or newlines
• maxResults - how many pages to process, up to 2000

Example input

{
  "urls": "https://example.com/, https://example.com/pricing, https://example.com/blog"
}

Use cases

• SEO audits - run a whole site list and get the issue table an audit report is built from
• Agency prospecting - audit a prospect's pages before the pitch and lead with concrete findings
• Pre-launch checks - verify titles, canonicals and indexability before a release goes live
• Migration validation - confirm that redirects land where they should and canonicals were updated
• Competitor analysis - see how competitors structure titles, schema and internal linking
• Content QA at scale - find thin pages, missing descriptions and duplicate H1s across hundreds of URLs

Which checks and why

The issues list applies the checks that actually affect search appearance and accessibility:

• Title 30 to 60 characters - shorter wastes the strongest ranking signal, longer gets truncated in results
• Meta description up to 160 characters - beyond that it is cut off; missing means the engine writes its own
• Exactly one H1 - zero leaves the page without a stated topic, several dilute it
• Canonical present - without it, parameter and duplicate URLs compete with each other
• noindex detection - the single most expensive accidental setting in SEO
• Viewport tag - its absence means the page is not treated as mobile friendly
• Image alt text - required for accessibility and the only signal image search has
• Thin content under 300 words - not a rule, but a reliable flag for pages that need work

Notes

• Pages are fetched as HTML and parsed server-side. Content injected by JavaScript after load is not seen, so single-page applications should be audited on their server-rendered output.
• Pages that cannot be fetched return a row with an error field instead of failing the whole run.
• schemaTypes collects every @type found in JSON-LD, including nested ones, which is the quickest way to see whether Product, Article, FAQ or Breadcrumb markup is present.
• wordCount is derived from the rendered body text including navigation, so compare pages within a site rather than across different templates.