Skip to main content
Help Center

Features

SEO AI Agent

Automatically crawl, audit, fix, and deploy SEO improvements to your live website.

What is the SEO Agent?

The SEO Agent crawls your website, identifies SEO issues, generates AI-powered fixes for you to review, and serves approved fixes to your live site through a lightweight script snippet.

How to Use

  1. Go to SEO Agent in the sidebar
  2. Select your website
  3. Choose a mode:
    • Audit: crawl and find issues only
    • Fix: find issues + generate proposed fixes
    • Deploy: same as Fix; approved fixes go live via the script snippet

What Gets Audited

Per page:

  • Missing, too-short, or too-long title tags
  • Duplicate title tags and meta descriptions across your pages
  • Meta description length issues
  • Missing canonical URLs
  • Missing Open Graph tags (og:title, og:description, og:image)
  • Missing JSON-LD schema markup (fixes are page-type aware: Article and FAQ pages get matching schema, everything else a safe WebPage)
  • Heading hierarchy problems (missing H1, with a suggested value, or multiple H1s)
  • Images without alt text (the audit names the offending images)

Site-wide:

  • AI-crawler access: whether your robots.txt blocks the agents that decide whether an engine can cite you (OAI-SearchBot, Claude-SearchBot, PerplexityBot, Bingbot, Google-Extended) and the training crawlers (GPTBot, ClaudeBot), which do not by themselves affect citation and are a separate decision, with copy-paste guidance
  • Sitemap presence and reachability
  • llms.txt presence (honestly labeled: an agent-readiness signal used by Anthropic/OpenAI agent tooling, not a Google ranking lever)
  • Core Web Vitals field data (LCP / INP / CLS at the 75th percentile, via Google's CrUX, only when real-traffic data exists)

Optional: IndexNow setup for faster Bing/Yandex recrawl after fixes (see below).

Two Ways to Apply Fixes

Apply at source (recommended)

Every approved fix shows a Copy HTML snippet button with ready-to-paste markup. Adding the fix directly to your site's HTML (or your CMS's SEO fields) is the strongest option: it's visible to every crawler, including the AI engines.

Why this matters: as of 2026, the major AI crawlers (GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, Meta-ExternalAgent) read your raw HTML only and do not run JavaScript. Google (and Gemini, which uses Google's rendering) does execute JavaScript, on a delayed second pass. Fixes applied at source reach all of them immediately.

Script snippet (quick patch, any website)

Add this to your site's <head>:

<script src="https://reachsurge.ai/seo-fix.js?key=YOUR_SITE_KEY" defer></script>

The script fetches your approved fixes and injects title, meta description, canonical, and schema markup into the rendered page.

Honest limitations of the script method:

  • Google picks the changes up via its JavaScript rendering (typically reflected in 2–8 weeks); most AI crawlers will not see script-applied fixes at all, because they don't execute JavaScript.
  • OG tag fixes via script won't be read by social media crawlers (Facebook, LinkedIn). Apply those at source.
  • Canonical is the weakest field to set via JavaScript (Google treats JS-set canonicals as a last resort). The agent only proposes a canonical when your page has none.

Use the script for an instant patch, then apply the same fixes at source when you can. Once a fix is in your source, roll back the script version from the fix history so the two never conflict.

Approval Workflow

All fixes are generated as "pending." You review and approve or reject each fix before anything goes live. Nothing is deployed without your approval.

Faster recrawl with IndexNow (optional)

On the SEO Agent page you can set up IndexNow to nudge Bing, Yandex, Naver and Seznam to recrawl a page as soon as you approve a fix. Google does not support IndexNow, so this has no effect on Google.

Setup is one-time and requires proving you own the domain: generate a key, host a small text file (named {key}.txt, containing only the key) at your site root, then click Verify. Until the file verifies, no pings are sent. After that, approving a fix pings automatically, and the weekly check re-pings pages whose fixes are confirmed live.

How long until I see results?

Applying a fix is the fast part. Search engines take time to recrawl and re-evaluate: Google typically reflects on-page changes in 2–8 weeks; AI engines recrawl on their own schedules. The fix status in your dashboard shows whether a fix is live on your pages. Rankings and citations move later.

Plan Requirements

  • Free: Audit mode only
  • Starter+: Full audit + fix + deploy

FAQ

How many pages does the agent crawl? Up to 20 internal pages per run, starting from your homepage.

Does it respect my robots.txt? Yes. The agent honors your robots.txt and crawls at a rate of 1 page every 2 seconds.

Can I roll back a deployed fix? Yes. Each fix stores the original value as a snapshot. Use the Roll back button in the fix history. The script stops serving that fix within about two minutes (a short cache window applies), and your page returns to its original values on the next load after that.

Previous

Knowledge Base

Next

White-Label Report Branding