Crawl, indexation, speed, and metadata diagnosed on your live URLs — findings ranked by the traffic they put at risk.
Run a technical SEO audit on metaflow.life. Focus on indexation and Core Web Vitals, and rank findings by impressions at risk.
Three critical issues account for most of the exposure: the blog canonical chain, 18 orphaned sitemap URLs, and an unpreloaded LCP image on the blog template. Fixing the canonical tag alone recovers 12 pages with existing impressions.
This technical SEO audit checklist runs as an agent workflow rather than a document you work through by hand. It crawls the site, cross-references what Google actually indexed, and returns each failure with the affected URLs, the likely cause, and the fix — so the first hour goes to decisions instead of data collection.
It is the ordered set of machine-side checks that decide whether your content can be found, fetched, and indexed at all: robots and sitemap correctness, crawl paths, canonical and duplicate handling, status codes, rendering, page speed, and metadata integrity. Everything else in SEO depends on these passing first.
It compares three sets — the URLs in your sitemaps, the URLs a crawl can reach, and the URLs Google reports as indexed. Anything in one set but not another is an indexation issue, and the report names which of the three lists each URL is missing from.
Below roughly ten thousand URLs it rarely limits anything, so the audit only raises crawl budget when it finds real waste: faceted parameter explosions, infinite calendars, redirect chains, or large blocks of soft 404s consuming fetches that should reach your money pages.
Field data first, because that is what Google scores, then the lab diagnosis that explains it. The audit flags which templates fail LCP, CLS, or INP thresholds and attributes the cause to a specific render-blocking resource, unsized element, or long task.
A full pass quarterly, plus a lightweight indexation and status-code check after every release or migration. Because the agent handles the data gathering, the post-release check takes minutes and catches the regressions that otherwise surface a month later as lost traffic.
The agent reads your site the way a crawler does and cross-checks against Google Search Console for impressions, coverage, and Core Web Vitals field data. PageSpeed Insights supplies lab diagnostics, and your CMS or repository provides the template context that turns a list of URLs into a list of fixes. Findings land in the run output and can be pushed to Slack, Linear, or a shared doc for whoever owns the fix.
Point the agent at the property and, if you only want part of the site audited, the path prefix to scope the crawl.
Decide what counts as urgent — for most teams that is anything blocking indexation of a page with existing impressions.
It fetches URLs, pulls coverage and field data, and groups failures by template rather than listing thousands of individual URLs.
Each item arrives with affected URLs, evidence, likely cause, and a proposed fix you can approve or reject.
Re-run scoped to the fixed template to confirm the issue cleared rather than auditing the whole site again.
Findings cite your URLs and your Search Console data, not generic best practice
Issues are grouped by page template, so one fix closes hundreds of URLs
Severity is ranked by impressions at risk instead of an arbitrary score
Re-runs are scoped, which makes verification cheap enough to actually do
Crawlability and robots directives, indexation coverage, status codes and redirects, canonical and duplicate handling, rendering, page speed, structured data validity, and metadata integrity. Content quality and link building are separate audits — keeping them apart is what makes each report short enough to act on.
It can describe what to check but cannot crawl your site or read your coverage reports, so its output is advice rather than diagnosis. This template runs inside an agent with crawl and Search Console access, which is the difference between "check your canonicals" and "these 41 URLs canonicalize to a noindexed page."
Minutes for a small site, longer for large crawls, and you get the report in one pass rather than assembling exports from three tools.
It consumes the same class of crawl data and adds the interpretation layer those tools leave to you — which findings matter, why, and in what order. Many teams keep a crawler for deep ad-hoc inspection and use this template for the recurring pass.
No. This template only diagnoses and proposes. Any fix ships through your normal deploy or CMS workflow after you approve it.