SEOBolt crawls your site for classic SEO and the new AI discovery layer: ChatGPT crawler access, llms.txt quality, entity schema, pricing context, and citation-ready pages. Then it hands your AI agent ready-to-apply fixes over MCP.
No signup and no payment details needed. Checks the homepage, robots.txt, sitemap.xml and llms.txt.
Connect your AI agent · MCP server and API on every plan, including Free
Download on theApp Store — iPhone, iPad & MacWe ran this exact loop on seobolt.app itself: an AI agent applied our fix plan and the health score went 67 → 93 in one session (83 pages, July 2026).
If ChatGPT Search, Claude, Perplexity, and other assistants cannot crawl or understand your product, you are invisible in a growing discovery channel. SEOBolt now scores that layer directly in audits, fix plans, API output, and MCP workflows.
Example output: illustrative numbers, not a real site
Overall
Google SEO
LLM Search
12
Missing titles, broken links
31
Thin metadata, missing alts
204
No issues found
Crawl your site for Google SEO issues plus ChatGPT Search, llms.txt, AI crawler, schema, and citation gaps
Every issue ships with the exact code snippet to fix it — pull the whole plan into Claude Code or Cursor via MCP and fix your site in one session
A daily or weekly email with each project's health score, its change since the last crawl, and error and warning counts
Scheduled re-crawls catch broken pages and missing titles; connect Search Console to see clicks, impressions, and average position per query and page
On Pro and Business, see the average position, clicks, and impressions Google reports for each query and page next to your audit. Not a live rank checker
Score key pages against a target keyword and see what to fix next
SEOBolt is designed to earn trust with real audit output, transparent data sources, and a useful free plan.
Run a real crawl
The free plan includes 2 projects and 3 crawls per month, so you can test the crawler on real pages before upgrading.
See every issue source
Audits show the affected URL, issue type, and fix guidance. Backlink and difficulty metrics explain how they are estimated.
Upgrade for workflow
The MCP server and API are on every plan. Paid plans are for recurring monitoring, more sites, larger crawls, and GSC data.
Start with a real crawl. The MCP server and API are on every plan. Upgrade when you need multi-page crawls, scheduled re-checks, fix-plan history, or client delivery.
No browser required. Sign up with one API call, connect via MCP, and let your AI assistant audit your site, pull the prioritized fix plan with code snippets, apply the fixes, and re-crawl to prove the score went up.
$ curl -X POST /api/agent-signup -d '{"name":"Claude"}'
{"api_key":"sk_live_...","project_id":"...","message":"Use this API key with the MCP server"}
Setup guides: Claude Code SEO audits, Cursor SEO audits, Codex CLI SEO audits, Claude Desktop SEO audits, and the Google Search Console integration. Then add the Claude Code SEO skills for a weekly SEO process. Coming from another tool? Compare the Ahrefs alternative, Semrush alternative, Screaming Frog alternative, or Moz alternative.
No black boxes. Here's exactly what each tool does and where the data comes from.
Our crawler visits every page on your site (up to 50K pages), parsing HTML with Cheerio. It checks technical SEO plus ChatGPT Search crawler access, llms.txt quality, AI entity schema, answer metadata, pricing context, and citation-ready copy.
Your website only. We respect robots.txt.
Connect Google Search Console (Pro and Business) to see the average position, clicks, and impressions Google reports for every query and page. SEOBolt does not run live Google rank checks.
Google Search Console API (read-only).
Keyword rankings from Search Console →We crawl candidate referring pages and record the links we find to your domain — source URL, anchor text, follow/nofollow status — then re-check that they are still live. Discovery through Google results needs a SERP data provider, so coverage is limited.
Page crawling of discovered sources. Not a proprietary index like Ahrefs (we're transparent about this).
A 0-100 authority score from Common Crawl's public web graph: each domain's harmonic-centrality rank is mapped to a log scale across the top 10 million domains. Domains outside that slice score 0, with the reason shown.
Common Crawl webgraph; imported backlinks only when a domain has no rank. Cached 7 days.
How domain rating works →We don't have Ahrefs' 15-year backlink index — we use Common Crawl and Search Console data. Our DR and KD are transparent formulas, not black boxes. What we do have: a fast crawler, clear audit reports, API/MCP access, and pricing built for recurring SEO operations.