SEO auditfor the age ofAI coding.
A Chrome side panel that audits every page the way a crawler sees it. Scores, findings, proactive JSON-LD suggestions — all copy-ready for Claude Code, Cursor, Copilot, and anything else on the other end of your keyboard.
The reading
LiveTally
severity- ErrorsBlocks indexing or ranking1
- WarningsSuboptimal — should fix1
- InfoSuggestion or hint1
- PassedCheck is healthy2
By category
6 areas- Meta tags62
- Heading hierarchy78
- Social previews92
- Structured data70
- Images85
- Indexing96
- Site-wide
Findings
5 totalAgent MCP
futureProgrammatic access for AI agents — MCP server. Let your coding agent run an audit and consume the results directly.
Six pillars.
What the crawler sees,
not what your DOM pretends.
The extension fetches the page's actual HTTP response and parses it with DOMParser — the same signal Google starts from. No mount-merging. No stale client routes. No false positives from SPA hydration.
- Real network response, not in-memory DOM
- Avoids CSR/SPA bleed between soft-nav routes
- Matches what a basic crawler ingests
1<!DOCTYPE html>2<html lang="en">3<head>4 <title>Build a React app in 2026</title>5 <meta name="description" content="A practical guide…" />6 <link rel="canonical" href="/react-2026" />7 <meta property="og:image" content="/og.png" />8 <script type="application/ld+json">9 {"@type": "Article", "author": …10 </script>11</head>12<body>13 <h1>Build a React app in 2026</h1>14 <p>React has evolved…</p>
Build a React app in 2026
Every JSON-LD block,
node by validated node.
Expandable tree of every structured-data block on the page. Schema.org types resolved, Google Rich Results specs checked field by field, unknown properties flagged inline. Copy a single value, a subtree, or the whole block.
- Per-node schema.org + rich-results validity
- Required vs. recommended field deltas
- Copy value · subtree · block
Not just validation.
What you're missing.
SEO Lens detects the kind of page it's looking at — article, product, homepage, category — and tells you which schema.org types you should add, with Google Rich Results specs and a minimal template to paste in.
- Page-kind detector (Article / Product / Home / …)
- Google Rich Results field validation
- Minimal JSON-LD templates, copy-ready
This page looks like an article. Adding Article structured data helps Google surface it in rich results.
Why we recommend this- ·Detected <article> element and byline
- ·Matches article URL pattern /blog/*
Template — paste into a <script type="application/ld+json">{ "@context": "https://schema.org", "@type": "Article", "headline": "", "image": [""], "datePublished": "", "author": { "@type": "Person", "name": "" } }
One page report.
Whole-site awareness.
robots.txt, sitemap.xml, security.txt, manifest, feeds, soft-404 probes, canonical host drift — folded into the current page's report so you catch the blocker that's hiding above the page level.
- robots.txt disallow detection
- Sitemap inclusion check
- Soft-404, canonical host drift, security.txt
Site-level signals
origin- sitemap/urlset · 2 · https://example.com/sitemap.xml
- https://example.com → 200
- https://www.example.com → 301 (↪)
- rss · https://example.com/rss.xml
The outline and the share card,
seen the way they're read.
Two views you actually use before shipping: the heading outline flattened into a tree so skip-levels jump out, and the Open Graph preview card rendered with your real hero image, title, and description.
- H1→H2→H3 tree with level badges
- Live OG card at 1200×630
- Real hero image, not a mock
- H1Introducing SEO Lens
- H3How it works
- H2Built for AI-assisted coding
- H3Copy-ready findings
- H3Proactive JSON-LD
- H2Roadmap
Four steps. One round-trip.
Open the side panel
Pin SEO Lens, click the icon. The audit runs against the active tab the moment the panel opens.
Watch the page get x-rayed
Fetched HTML is parsed. 30+ rules run. Score, categories, and findings appear in under 150ms.
Copy any finding
Every finding has its own copy button — structured as plain text with rule, severity, and snippet.
Paste into your agent
Hand it to Claude Code, Cursor, or Copilot. The model has everything it needs to ship the fix.
Seven categories. Thirty rules.
The audit engine lives in an Effect-TS package, pure and runtime- agnostic. Today it powers the extension. Tomorrow it powers the CLI, the MCP server, and anything else we (or you) plug it into.
Title & description
- Title length
- Meta description length
- Title present
Structure & hierarchy
- Single H1 enforcement
- Skip-level detection
- Heading tree
OG & Twitter cards
- og:title / og:image
- Twitter card type
- Image URL validity
JSON-LD & Rich Results
- schema.org type recognition
- Google Rich Results fields
- Proactive type recommendations
Alt text & accessibility
- Missing alt detection
- Empty alt warnings
- Image inventory
Canonical & robots
- Canonical URL
- robots meta directives
- noindex / nofollow
Cross-page signals
- robots.txt disallow
- Sitemap inclusion
- Soft-404, canonical host drift
- security.txt, manifest, feeds
Site-wide audit — crawl the full sitemap, score every page.
More surfaces — CLI, MCP server, and beyond.
Install it. Open the panel. Ship.
Free. Manifest V3. No account. Chrome now, Firefox shortly, Safari & Edge after.